Plant phenotyping methods.

CITRUS FLOWER, FRUIT, AND SHOOT RECOGNITION BASED ON IMPROVED YOLOv10 →

In the process of agricultural intelligence, precise detection of plant organs serves as the foundation for core tasks such as crop phenotyping analysis and yield prediction. However, in complex field environments, small targets such as citrus flowers and shoots face challenges including scale variation, background interference, and dense occlusion, which severely impact detection accuracy. This study improves the YOLOv10 model by introducing the BAM (Bottleneck Attention Module) attention mechanism and GIoU (Generalized Intersection over Union) loss function, constructing a YOLOv10s-BAM-GIoU model suitable for citrus flower, fruit, and shoot recognition. The BAM attention mechanism enhances the model's feature extraction capability for small target organs under complex backgrounds through parallel channel and spatial attention branches; the GIoU loss function improves the localization accuracy of densely occluded targets by optimizing the geometric alignment between predicted and ground-truth boxes. Validation experiments were conducted on a self-constructed dataset. The experimental results show that the improved YOLOv10s achieves significant advantages in comprehensive detection accuracy, with an mAP50 of 89.1%, representing an improvement of 2.9%~9.5% over the original YOLOv10s and other comparative models. In fine-grained category detection, the model achieves mAP50 of 91.2%, 83.6%, and 92.5% for shoots, flowers, and fruits, respectively. Furthermore, while maintaining high detection accuracy, the model achieves a detection speed of 23.6 ms per frame, meeting real-time detection requirements. The research results demonstrate that the improved YOLOv10s model integrating the BAM attention mechanism and GIoU loss function achieves an optimal balance between accuracy and speed in citrus organ detection tasks, providing a preferred solution for field real-time detection systems.

Why it matches plant phenotyping methods柑橘の花・果実・シュートという植物器官を画像から検出する改良モデルを開発し、データセットで精度と速度を検証しており、表現型取得手法が中心である。

abstractThis study improves the YOLOv10 model by introducing the BAM (Bottleneck Attention Module) attention mechanism and GIoU (Generalized Intersection over Union) loss function, constructing a YOLOv10s-BAM-GIoU model suitable for citrus flower, fruit, and shoot recognition.

From detection accuracy to safety assurance in intelligent plant health early warning systems →

Plant disease and plant stress early warning systems have advanced through deep learning, remote sensing, digital phenotyping, disease forecasting, and sensor networks. Detection accuracy, precision, recall, F1-score, and area under the curve remain indispensable, but they are insufficient for judging whether a warning can support timely and proportionate phytoprotection under field variability. This Mini Review argues that intelligent plant health warning systems should be evaluated not only as prediction models, but also as safety-relevant decision-support systems embedded in biological, agronomic, and operational contexts. We first relate AI-based detection to established plant disease forecasting and decision-support traditions, including weather-based models, epidemiological forecasting, and integrated disease management. We then adapt selected safety-assurance concepts, including risk assessment, failure mode and effects analysis, Bow-tie reasoning, warning-threshold governance, reliability analysis, resilience thinking, and response closure, to host-pathogen-environment warning chains. The proposed framework links AI or sensor outputs with pathogen biology, host susceptibility, environmental conduciveness, inoculum pressure, uncertainty assessment, risk classification, threshold decisions, human or automated verification, intervention, and feedback learning. Illustrative crop-pathogen scenarios, including wheat rust, rice blast, potato late blight, grapevine downy mildew, and citrus greening, show how safety assurance can complement existing forecasting and decision-support systems rather than replace them. The framework remains conceptual, and whether these added assurance functions improve existing warning systems requires comparative evaluation under field conditions. Future systems should be evaluated through detection performance and response-oriented indicators such as lead time, calibration, false-alert burden, missed-warning rate, response completion, disease suppression, economic value, and learning after field action.

Why it matches plant phenotyping methods植物病害・ストレスの検出を含む知的警戒システムについて、AI・リモートセンシング・デジタルフェノタイピング・センサーネットワークの評価枠組みを体系的に論じる方法論レビューであり、方法論が中心です。

abstractPlant disease and plant stress early warning systems have advanced through deep learning, remote sensing, digital phenotyping, disease forecasting, and sensor networks.

YOLO-Based Deep Learning for Citrus Fruit Detection, Counting, and Yield Estimation in Complex Orchard Environments: A Systematic Review →

Abstract A systematic review based on 174 Scopus-records of studies using YOLO-type one-stage detectors for detecting citrus fruits, their count, and yield estimation followed PRISMA guidelines 2020. The title/abstract-screening process, done in duplicate (κ=0.920) yielded 90 included study-records, followed by two further post-hoc exclusions. Each study in the 80 reporting on fruit-level detection showed an average precision of 89.6%, recall of 85.9%, and mAP@0.5 of 91.0%. However, coverage for any individual metric rarely exceeded half of the studies, and only 13% were able to report the more stringent mAP@0.5:0.95. Both YOLOv8 and YOLOv5 were each utilized as the backbone architecture by approximately 22.2% of the studies. From 2025, YOLOv11 has also been emerging. Half of all studies modified architectural components including attention modules, lightweight architectures, and variants of IoU loss functions. Original contributions are generally concentrated in downstream tracking, sensor fusion, and yield modeling rather than the detector itself. A custom-made seven-domain risk of bias tool was developed and utilized by two reviewers who arbitrated discrepancies (91.5%). Results showed that all but one of the reviewed studies had a high level of risk due to almost universal lack of statistical validation and limited dataset diversity; a sensitivity analysis excluding the most risky studies left the performance profiles nearly identical. We conclude that the field has converged around a common technical toolkit but continues to lack standardized benchmarks, multispectral data, and rigorous field-deployment validation.

Why it matches plant phenotyping methods柑橘果実の検出・計数・収量推定に用いる画像解析手法を体系的にレビューし、性能評価、リスク・オブ・バイアス、標準化やベンチマーク不足を検討しており、植物フェノタイピング手法が中心である。

abstractA systematic review based on 174 Scopus-records of studies using YOLO-type one-stage detectors for detecting citrus fruits, their count, and yield estimation followed PRISMA guidelines 2020.

Identification of spectral biomarkers for early fungal decay in navel oranges by Vis-NIR hyperspectral imaging and multi-scale feature fusion. →

Early detection of latent fungal decay caused by Penicillium italicum(P. italicum) and Penicillium digitatum(P. digitatum) remains challenging due to the absence of visible symptoms. In this study, a Vis-NIR hyperspectral imaging framework was developed to characterize early biochemical alterations in navel oranges. To address sample scarcity, a generative modeling approach (WGAN-GP) was employed to capture the intrinsic physiological variability of infected tissues. The successive projections algorithm (SPA) identified 20 key wavelengths associated with water redistribution (OH), carbohydrate depletion (CH), and chlorophyll degradation. These wavelengths were expanded into continuous ROI windows (W = 17), enabling integration of narrow-band pigment signals and broad-band absorptions related to water and carbohydrates via a multi-scale mixture-of-experts (MS-MoE) network. The framework achieved a classification accuracy of 97.10% and an F1-score of 0.9666. These results demonstrate that specific spectral absorption windows can serve as reliable, chemically interpretable spectral biomarkers for detecting early pathological changes in citrus fruit.

Why it matches plant phenotyping methodsVis-NIRハイパースペクトル画像と解析モデルを開発し、柑橘果実の初期病変をスペクトル特徴から推定する方法が研究の中心であるため、植物病害表現型の計測手法として含める。

abstracta Vis-NIR hyperspectral imaging framework was developed to characterize early biochemical alterations in navel oranges.

Paper-based nanozyme sensor array for volatile chemical fingerprinting of Huanglongbing-associated citrus samples. →

Citrus Huanglongbing (HLB) is one of the most destructive citrus diseases worldwide, and early diagnosis remains challenging because uneven pathogen distribution often leads to false-negative PCR results. Here, a paper-based nanozyme sensor array was developed for volatile organic compound (VOC) chemical fingerprinting of HLB-associated citrus samples. Fe/Al bimetallic NH₂-MIL-53 nanozymes were designed to regulate VOC adsorption and peroxidase-like catalytic activity. Exposure of the nanozyme sensors to VOCs reduced the catalytic oxidation of 3,3',5,5'-tetramethylbenzidine (TMB), generating concentration-dependent colorimetric responses. By tuning the Fe/Al ratio, the nanozymes exhibited differentiated responses toward HLB-associated volatiles, including methyl salicylate, phenylacetaldehyde, and linalool, with recognition limits of 0.1-0.5 ppm. Integration with two MOF-based sensing channels formed a five-channel artificial olfactory array capable of generating multidimensional color fingerprints. The sensor array successfully discriminated healthy, asymptomatic, and infected citrus leaf samples, achieving an overall classification accuracy of 92.5%. These results suggest that the proposed platform provides a simple and low-cost approach for citrus VOC chemical fingerprinting and may provide useful information for quality-related screening in citrus production systems, while further field-oriented validation is still needed to assess its practical applicability.

Why it matches plant phenotyping methods柑橘葉のHLB感染状態をVOCセンサーアレイで識別する計測・解析法の開発が研究の中心であり、植物病害状態を直接推定している。

abstractHere, a paper-based nanozyme sensor array was developed for volatile organic compound (VOC) chemical fingerprinting of HLB-associated citrus samples.

Deep learning enables edge deployment for citrus leaf disease recognition in natural orchard scenes. →

Background Accurate and efficient detection of citrus leaf diseases is important for orchard monitoring, early intervention, and intelligent disease management. However, practical application in natural orchard environments remains challenging because of complex backgrounds, large variation in symptom scale, strong interclass similarity, and limited edge computing resources. Methods In this study, a lightweight object detection model, SKM-YOLOv11, was developed based on YOLOv11n for citrus leaf disease recognition in natural scenes. A self-built dataset was constructed from public images and field images, containing 6414 images across six classes. StarNet-S050 was introduced as the backbone, C3k2-Star was designed to enhance feature fusion across scales, and a lightweight shared detection head, MNS-Head, was constructed to reduce prediction redundancy. Results Compared with YOLOv11n, SKM-YOLOv11 increased Precision, Recall, and mAP @0.5 by 2.92, 1.47, and 0.95 percentage points, respectively. Meanwhile, FLOPs, parameter count, and model size were reduced by 31.7%, 32.7%, and 34.55%, respectively. Edge deployment on Jetson Orin NX Super achieved an inference speed of 72.45 frames per second. Conclusion The proposed model has strong potential for real-time citrus disease screening on resource-constrained devices and provides a feasible solution for edge-based intelligent disease management in natural orchards.

Why it matches plant phenotyping methods柑橘葉の病害状態を画像から認識する軽量物体検出モデルを開発し、データセット、精度比較、エッジ実装性能まで評価しており、植物表現型（病害状態）の取得・推定手法が中心である。

abstracta lightweight object detection model, SKM-YOLOv11, was developed based on YOLOv11n for citrus leaf disease recognition in natural scenes.

Linking plant water status dynamics to yield and fruit cracking in citrus orchards using UAV multi-sensor data and machine learning →

Citrus fruit cracking causes substantial yield and economic losses, yet its relationship with plant water status (PWS) and irrigation management remains insufficiently characterized. Unlike previous UAV-based irrigation studies that focused on water-stress detection or yield estimation, this study introduces a dynamic, physiology-based framework that links temporal PWS trajectories during key phenological stages to fruit-cracking risk at the individual-tree scale. UAV-based multispectral, thermal, and LiDAR data, combined with field physiological measurements and machine-learning models, were evaluated in an irrigation management experiment in an ‘Ori’ mandarin orchard (Israel) across three contrasting growing seasons (2023–2025). Several irrigation treatments with different irrigation timings and water inputs were applied during the growing season to evaluate their effects on temporal PWS dynamics and fruit cracking. Trunk growth (TG), stem water potential (SWP), stomatal conductance (SC), and plant area index (PAI) were measured throughout the two seasons and estimated using Random Forest models (R 2 > 0.783). These indicators were subsequently used to predict yield and fruit cracking with high accuracy (yield: R² = 0.896; cracking: R² = 0.845). Cracking was lowest in 2023 (∼3%), with ∼25% lower irrigation, suggesting reduced irrigation may reduce cracking risk. Higher cracking in 2024 (∼14%, vs ∼8% in 2025) coincided with intense heat events. Mid-season SWP and SC were strongly associated with yield formation and cracking patterns. These findings demonstrate that monitoring temporal PWS dynamics can support precision irrigation management by identifying high-risk zones and enabling irrigation strategies that stabilize PWS, reduce the incidence of cracking, and improve yield under variable climatic conditions.

Why it matches plant phenotyping methodsUAVマルチセンサーと機械学習により、樹体水分状態などの植物形質を推定し、収量・果実裂果を予測する技術的枠組みが研究の中心である。

abstractthis study introduces a dynamic, physiology-based framework that links temporal PWS trajectories during key phenological stages to fruit-cracking risk at the individual-tree scale.

Non-Destructive Prediction of Soluble Solid Content in Kumquats Using a Multi-Scale Convolutional Neural Network →

Traditional methods for detecting the soluble solid content (SSC) of kumquats are often destructive, time-consuming, and inefficient. In this study, a multi-scale convolutional neural network (MS-CNN)-based method is proposed for the rapid and non-destructive prediction of kumquat SSC. By integrating near-infrared spectroscopy (900–1700 nm) with deep learning, 424 spectral samples of kumquats were collected and modeled using the MS-CNN framework. The proposed model adopts a multi-scale feature extraction structure inspired by the Inception architecture, which effectively enhances the representation of spectral features and reduces overfitting. Experimental results showed that the MS-CNN achieved an Rp2 of 0.88, an RMSEP of 0.62 °Brix, and an MAEP of 0.51 °Brix on the internal prediction set. Among the evaluated models, the MS-CNN achieved the highest Rp2, while its RMSEP was comparable to that of PLSR and lower than those of SVR, BP, CNN, and BiLSTM. The proposed approach enables fast, accurate, and non-destructive prediction of kumquat SSC, providing a novel technical solution for fruit quality assessment. This work holds significant theoretical and practical value, and future efforts will focus on expanding the dataset, optimizing the network structure, exploring multi-index joint prediction, and promoting its real-world application.

Why it matches plant phenotyping methodsカンキツ果実のSSCという植物器官形質を、近赤外分光とMS-CNNで非破壊推定する手法の開発・比較評価が研究の中心である。

abstracta multi-scale convolutional neural network (MS-CNN)-based method is proposed for the rapid and non-destructive prediction of kumquat SSC

Design and Development of a Deep Learning-Based System for Multi-Fruit Disease Classification and Severity Detection Using VGG-16 and VGG-19 Architectures on an Expert-Verified Indian Fruit Crop Dataset →

Fruit diseases caused by fungal, bacterial, and viral pathogens cause devastating pre- and post-harvest losses to Indian agriculture, particularly in Maharashtra where Guava, Mango, Orange, Papaya, and Pomegranate are major horticulture crops. Accurate, early-stage disease identification directly impacts farmer income, food security, and precision crop management. Conventional manual inspection by agronomists is subjective, time-consuming, and not scalable across thousands of orchard acres. Automated deep learning-based image analysis has emerged as a transformative solution, offering high accuracy, speed, and field deployability. Existing deep learning models for plant disease detection are predominantly trained on the real time database, which inadequately represents Indian fruit crop species. Furthermore, published systems focus on binary disease presence detection and lack disease severity grading — a critical requirement for treatment decision-making. The absence of expert-verified datasets and the visual overlap among disease classes (such as Phytophthora vs. Scab in Guava, and Powdery Mildew vs. Ring-spot in Papaya) present significant classification challenges. Limited training data for rare classes such as Stylerandroot (Guava, 310 samples) and Bacterial Blight (Pomegranate, 304 samples) further compounds model generalization. This study presents an end-to-end deep learning framework using VGG-16 and VGG-19 architectures trained on a novel, expert-verified dataset of 7,372 approved images spanning 22 disease/healthy classes across 5 fruit types, after on filed validation. The pipeline includes image preprocessing (CLAHE, Gaussian denoising, normalization), hybrid multi-feature extraction (CNN features, GLCM, LBP, Color Histograms), transfer learning with progressive fine-tuning, and a cascaded severity estimation module. Three algorithms are designed: (1) a Transfer Learning Classification Algorithm using VGG-16/VGG-19 backbone with softmax multi-class head, (2) a Hybrid Feature Fusion Algorithm combining CNN deep features with handcrafted descriptors for improved minority-class performance, and (3) a Cascaded Rule-CNN Severity Estimation Algorithm classifying disease progression into Healthy, Mild, Moderate, and Severe categories. VGG-19 achieved 96.1% overall accuracy, 95.1% precision, 94.5% recall, and a macro F1-score of 0.942, significantly outperforming VGG-16 (94.5% accuracy, F1: 0.918). Mango classification achieved the highest accuracy at 97.8%, while severity estimation reached 91.2% overall accuracy with the Mild category being the most challenging at 86.4%.The proposed system demonstrates that expert-verified, domain-specific datasets combined with transfer learning and hybrid feature fusion significantly advance the state of fruit disease detection for Indian agriculture. This framework provides a scalable, interpretable, and practically deployable solution for precision horticulture.

Why it matches plant phenotyping methods植物画像から病害状態と重症度を推定する深層学習手法、データセット、評価を中心的に開発しており、植物表現型計測の方法論的研究に該当する。

abstractThis study presents an end-to-end deep learning framework using VGG-16 and VGG-19 architectures trained on a novel, expert-verified dataset of 7,372 approved images spanning 22 disease/healthy classes across 5 fruit types

Integrated assessment of drought-driven vegetation declines in Olive and Citrus Orchards of Semi-Arid Morocco: A Multi-Index Remote Sensing Framework →

Abstract In the era of climate change, drought is defined as one of the most severe natural catastrophes that affects the environment, crop growth, and water resources, leading to economic losses, migration, and risks to human life. Since 2019, Morocco has suffered one of the most severe droughts in its recording history, coinciding with a broader period of precipitation deficit across the Mediterranean basin, resulting in a significant reduction in reservoir storage levels and the suspension of irrigation provided by dams in some areas due to low or absent rainfall, making drought a serious challenge to natural resources in this country. This study focused on semi-arid regions, especially on the Tensift basin in Morocco, and was conducted between 2018 and 2024. The effects of drought on arboriculture were analyzed thanks to remote sensing by using the normalized difference vegetation index (NDVI), while NDVI, TCI (temperature condition index), VCI (vegetation condition index), VHI (vegetation health index), and SPI (standardized precipitation index) were employed to assess and evaluate the health of the vegetation area, especially the arboriculture area, and the effective water stress conditions in our agricultural study area. The analyses indicated that arboriculture cover decreased markedly, from 11.17% in 2020 to 7.40% in 2023. From the analyses, it was evident that the cover of this culture had significantly declined from 11.17% in 2020 to 7.40% in 2023. From 2018 to 2024, more than half of the area suffered from drought in the agriculture of varying intensity levels, from moderate to severe. The meteorology of the evaluations confirmed the above observations by showing that there was a considerable decline in precipitation levels from about 350 mm in 2019 to less than 50 mm in 2024, coupled with continuously negative SPI-6 indices. Moreover, it was established that there was a significant effect on tree crops such as olives and citrus. Degradation of land was at its worst during 2021 when it affected more than 3,000 hectares of olives and 2,250 hectares of citrus. Similarly, 80% of farmers engaged in the production of citrus registered a decline in yield from 37% to 41%. Consequently, this study provides new information concerning the need for comprehensive evaluation and monitoring of agricultural drought in Morocco, thereby emphasizing the importance of essential factors.

Why it matches plant phenotyping methodsリモートセンシングと複数の植生・水ストレス指数を中核に、オリーブ・柑橘樹の植生健康状態や干ばつ影響を評価しており、植物状態の抽出手法の実質的応用に該当する。

abstractThe effects of drought on arboriculture were analyzed thanks to remote sensing by using the normalized difference vegetation index (NDVI), while NDVI, TCI (temperature condition index), VCI (vegetation condition index), VHI (vegetation health index), and SPI (standardized precipitation index) were employed to assess and evaluate the health of the vegetation area, especially the arboriculture area, and the effective water stress conditions in our agricultural study area.

Editorial: Plant phenotyping for agriculture →

Modern agriculture operates at an unprecedented crossroads, it must simultaneously accelerate crop yields to feed an expanding global population and adapt to the severe, fluctuating pressures of climate change, structural soil degradation, abiotic water deficits, and evolving biological threats. Historically, selecting resilient crop varieties and implementing field-scale management strategies relied extensively on destructive, labor-intensive, and fundamentally subjective visual metrics. This manual processing approach has long been recognized as the primary operational bottleneck in agricultural advancement.To bridge the gap between rapidly expanding genomic data and actual field performance, the systematic, non-destructive quantification of structural and functional plant traits, plant phenotyping, has emerged as a transformative frontier. By integrating high-throughput engineering, multi-scale remote sensing, deep learning, and advanced molecular biology, modern phenotyping transitions crop science away from qualitative estimation toward highly reproducible, multidimensional data frameworks. This Research Topic presents new advances in advanced 3D reconstruction and deep semantic segmentation at the seedling stage; amodal fruit segmentation, morphological extraction, and early water-stress diagnostics; high-throughput in-field seedling counting and dynamic density modeling; multimodal foundation models, network pruning, and intelligent phytoprotection; aerial and spaceborne remote sensing for canopy analysis and weed monitoring; plant physiology, functional spectroscopy, and functional genomics under abiotic stress; and automated diagnostics for real-time orchard scouting and vineyard management.Automating the characterization of complex spatial layouts under controlled or greenhouse environments is essential for early variety selection and early-stage structural evaluation. Several contributions within this volume provide key breakthroughs in navigating overlapping tissues, severe occlusions, and low-contrast edge regions. showcases how substituting standard convolutions with deformable convolutions enables deep neural networks to accurately isolate the main stem of mature, high-density crops like soybeans. This architecture overcomes the traditional challenges of color mimicry and severe occlusion by pods and leaves, achieving an outstanding mIoU of 90.58% and providing reliable indices for lodging resistance and structural yield modeling (R 2 = 0.9746).Accurately extracting fruit morphology under commercial greenhouse conditions remains heavily constrained by overlapping crop structures, foliage cover, and variable shadows. Simple semantic masks typically fail when a target fruit is partially blocked, leading to a loss of key volumetric data.To resolve the challenge of hidden boundaries, Li, Yin, et al. (2025) developed CGA-ASNet, a specialized RGB-D amodal segmentation network driven by a Contextual and Global Attention (CGA) module designed to restore occluded tomato regions. Trained on a high-fidelity synthetic greenhouse dataset (Tomato-sim) generated via NVIDIA Isaac Sim's Replicator Composer and optimized with a mean coordinate fusion algorithm for real-world validation, this architecture expands the network's receptive field to predict the complete, hidden circular forms of occluded tomatoes, achieving an F@0.75 score of 94.2 and an amodal mIoU of 82.4%. This proves that simulation-to-real (Sim2Real) domain pathways can successfully decode full physical volumes under dense commercial canopies.Complementing this structural restoration, Yang, Li, et al. (2025) designed an integrated diagnostic framework to identify early water stress dynamics in greenhouse tomatoes. Built upon an optimized YOLOv11n core, their system integrates adaptive kernel convolutions (AKConv) into the network backbone's C3k2 modules and implements a recalibration feature pyramid detection head based on the specialized P2 small-target layer. This combination achieved a 5.4% increase in mAP50-95 for identifying fine phenotypic parts. By applying automated geometric analysis to the extracted bounding boxes, the system extracts plant heights and petiole count with low relative errors, feeding these phenotypic parameters into a Random Forest classification routine that flags water-stressed plants with 98% accuracy to guide targeted, automated drip irrigation.Accurate plant stands during early vegetative stages represent the foundational metric required to establish true field emergence rates, validate seed vigor across diverse breeding blocks, and perform early yield predictions.To solve the challenges of small targets, extreme spatial density, and adjacent leaf overlap, Zang et al. (2025) designed DM_IOC_fpn, a wheat seedling counting framework that balances local and global contextual features. By structuring a point-annotated dataset and embedding a densityenhanced encoder module, their network balances micro-scale spatial limits with macro-scale canopy structures. Optimized through a combined loss function tracking counting, classification, and regression parameters, this architecture achieved low error scores (RMSE = 2.91; MAE = 2.23), outperforming standard object-detection benchmarks in complex field environments.At the same time, scaling up to real-time aerial monitoring required major reductions in model complexity to support resource-constrained edge computers on autonomous aerial platforms. Feng, Nie, and Li (2025) engineered an ultra-lightweight YOLOv8n variant tailored for real-time maize seedling counting from high-speed UAV RGB overflights. By reparametrizing RepConv with HGNetV2, they constructed a lean Rep_HGNetV2 backbone, integrated a Bidirectional Feature Pyramid Network (BiFPN) for multi-scale feature alignment, and implemented a Task Dynamically Aligned Detection Head (TDADH). This architecture compressed total model parameters by 47% and reduced weight sizes to 3.5 MB while maintaining a 96.5% detection accuracy and an ultra-fast processing speed of 146.3 FPS, paving the way for low-cost, real-time field scouting.Automated phytoprotection requires machine-vision architectures capable of generalizing across highly diverse species, complex field conditions, and varying computational boundaries. A significant subset of the published papers addresses these challenges through foundation model adaptation, multi-modal alignment, and efficient network compression.A major paradigm shift presented in this collection involves moving away from task-specific training and toward foundation model adaptation. Chen, Ruan, et al. (2026) introduce a novel architecture integrating the DinoV3 foundation model with a Unet framework to achieve robust leaf lesion segmentation across diverse species (such as coffee and black gram). By incorporating a Spatial Prior Module (SPM), their approach surpassed standard benchmark networks by over 10.5% in IoU while reducing inference times by approximately 93.6%, demonstrating that highparameter foundation models can be highly optimized for resource-constrained edge devices in real-time scouting.To solve the perennial problem of limited training data for rare or emerging crop diseases, Cooper et al. ( 2026) developed an ingenious synthetic data generation pipeline. Combining 3D procedural leaf modeling in Blender with diffusion-based disease synthesis (Stable Diffusion fine-tuned with LoRA and ControlNet), they synthesized highly accurate plant disease images with perfect groundtruth annotation masks. When deployed in low-resource data settings, combining these synthetic pipelines with restricted real-world datasets consistently drives significant improvements in downstream segmentation tasks. To tackle specific, complex pathologies, Xu, Chang, et al. (2025) developed the TSSC deep learning model, which embeds three-neighbor channel attention paired with a complementary squeeze-and-excitation mechanism. This specific architecture minimizes structural degradation risks while pushing classification accuracy to 99.61% for highly complex pea leaf pathologies. Similarly, Feng, Liu, et al. (2025) tackled overlapping leaf occlusions and small lesion footprints in citrus groves with YOLO-Citrus, an optimized framework integrating C3K2-STA, ADown modules, and a Wise-Inner-MPDIoU loss function to strike a balance between edge computational constraints and field deployment.UAVs and high-resolution satellite imagery have expanded the operational scale of phenotyping from individual pots to vast breeding blocks and commercial fields, allowing researchers to capture macro-dynamic parameters over time.In complex canopy systems that defy standard top-down aerial sensing, such as single-staked white Guinea yams, Iseki et al. (2026) demonstrated the distinct advantage of utilizing multi-angle (combined nadir and oblique) UAV imaging configurations. When coupled with support vector regression, this method captures complementary canopy-structure information to model shoot biomass trajectories (R 2 = 0.79) across multiple years and management zones. These nondestructive, time-series datasets enabled the fitting of genotype-specific Richard's growth curves using Bayesian inference, isolating valuable genetic variations in early growth allocation.To capture full-season vertical physiological changes over large scales, Li, Yue, and Luo (2025) developed a hybrid CNN-LSTM-Attention (CLA) model designed to estimate the full-period Leaf Area Index (LAI) in rice using multi-temporal UAV multispectral imagery. By using the CNN layer to extract instantaneous spatial features, the LSTM block to process seasonal time-series intervals, and a self-attention mechanism to weight critical growth transitions, their platform achieved a high coefficient of determination (R 2 = 0.92) and kept relative root mean square errors (RRMSE) below 9%. This network minimized soil background noise during early vegetative stages (LAI values 1-

Why it matches plant phenotyping methods植物フェノタイピングの技術動向を扱うEditorialであり、画像解析、UAVセンシング、深層学習、形質抽出などの方法が中心的に整理されている。

titleEditorial: Plant phenotyping for agriculture

Deep Learning and Computer Vision for Crop Maturity Assessment →

Maturity at harvest is a critical determinant of yield, storability, market value, and nutritional quality, making accurate and objective maturity assessment essential for sustainable crop and fruit production. The agricultural sector is under pressure to satisfy rising global food demand while reducing losses and environmental impacts, yet conventional maturity assessment methods remain largely manual, subjective, and labour-intensive. Against this backdrop, computer vision and deep learning have emerged as powerful tools for non-destructive, high-throughput evaluation of maturity traits in the field and along the supply chain.​ This review consolidates recent advances in deep learning-based maturity assessment across a wide range of crops, with a particular emphasis on citrus fruits, where external colour change, internal quality, and heterogeneous orchard conditions pose distinctive challenges. The paper analyzes state of the art architectures for classification, segmentation and detection, associated datasets and imaging modalities, and the metrics used to benchmark performance. By critically examining their advantages and limitations for real-world deployment, the review outlines key research gaps and future directions toward robust, scalable, and sustainable DL-driven maturity assessment systems for both citrus and other major crops.

Why it matches plant phenotyping methods作物の成熟度という植物形質を対象に、画像・深層学習による評価手法、データセット、画像モダリティ、ベンチマーク指標を体系的にレビューしており、フェノタイピング手法が中心です。

abstractThis review consolidates recent advances in deep learning-based maturity assessment across a wide range of crops

External-field optimality of logarithmic coeffcients with applications to geometric image analysis →

This paper develops an external-field optimization framework for logarithmic coefficients of multiplier-defined univalent functions and establishes optimality principles with applications to geometric image analysis. Using the Herglotz representation, logarithmic coefficients are expressed as nonlinear moment functionals of probability measures on the unit circle, where multiplier coefficients act as an external field imposing admissibility constraints. A general Euler–Lagrange condition is derived, yielding a nonlinear equilibrium equation that characterizes extremal solutions. When the external-field potential admits a unique maximizer, the extremal measure collapses to a one-point distribution, leading to explicit optimal special-function solutions. The theoretical framework is applied to a dataset of citrus lesion images. After segmentation and conformal normalization of lesion regions, approximate logarithmic coefficients are computed from boundary harmonic expansions. A distortion index and harmonic separation criterion are introduced, and a coefficient separation theorem is verified numerically, demonstrating that geometric differences in lesion morphology correspond to measurable differences in logarithmic coefficient distributions. The results provide a mathematically rigorous connection between nonlinear external-field optimization, conformal special-function representations, and shape-based image descriptors. This approach offers a theoretically grounded and conformally invariant methodology for analyzing geometric irregularity in biomedical and agricultural imaging.

Why it matches plant phenotyping methods柑橘病斑画像を分割・正規化し、病斑形態を定量化する新しい画像記述子と判定基準を導入・検証しており、植物の病害状態の表現型抽出が中心である。

abstractThe theoretical framework is applied to a dataset of citrus lesion images. After segmentation and conformal normalization of lesion regions, approximate logarithmic coefficients are computed from boundary harmonic expansions.

Klasifikasi Penyakit Tanaman Jeruk Berdasarkan Citra Daun Menggunakan Metode Convolutional Neural Network Arsitektur EfficientNetV2-S →

The classification of citrus leaf diseases still largely relies on traditional assessment by farmers, which may lead to errors in identifying disease types. Previous studies have widely applied Convolutional Neural Networks (CNNs) for plant disease classification; however, most have utilized first-generation EfficientNet architectures, while the application of EfficientNetV2-S for citrus leaf disease classification remains relatively limited. Furthermore, the implementation of a progressive fine-tuning strategy on the EfficientNetV2-S architecture for this task has not been extensively investigated. Therefore, this study aims to implement the EfficientNetV2-S architecture for citrus leaf disease classification. The dataset used was the Citrus Leaves Prepared dataset from Kaggle, consisting of 596 images categorized into four classes: blackspot, canker, greening, and healthy. The data underwent preprocessing and image augmentation, including flipping, rotation, and zooming, before being divided into training, validation, and testing sets with a ratio of 70:10:20. The model was developed using a transfer learning approach combined with progressive fine-tuning. Experimental results demonstrated that the proposed model achieved a testing accuracy of 93.33% under the 100-epoch training scenario. With this level of accuracy, the model shows strong potential for implementation as an early detection system for citrus leaf diseases, assisting farmers in making timely and appropriate decisions to prevent crop failure.

Why it matches plant phenotyping methods柑橘葉の病徴を画像から分類するCNN手法の実装・評価が研究の中心であり、植物の病害状態を直接推定するため採用。

abstractExperimental results demonstrated that the proposed model achieved a testing accuracy of 93.33% under the 100-epoch training scenario.

Finding Hidden Huanglongbing using an Electronic Nose. →

Background and Objective: The Huanglongbing (HLB) is one of the most destructive diseases affecting citrus worldwide. A major challenge in its management is its ability to remain asymptomatic for extended periods, delaying timely detection and control. This study aims to develop and evaluate a compact electronic nose (e-nose) system equipped with metal-oxide semiconductor (MOS) sensors for the early detection of Candidatus Liberibacter asiaticus infection in citrus leaves through Volatile Organic Compound (VOC) analysis under field-like conditions. Materials and Methods: A total of 454 Purworejo Siamese citrus leaf samples were collected from two orchards. The infection status of each sample was confirmed using conventional Polymerase Chain Reaction (PCR) prior to headspace VOC extraction. The cross-sensitive MOS sensor array converted VOC interactions into electrical signals, which were subsequently preprocessed, feature-extracted and analyzed using machine learning pipelines. Model selection and optimization were performed on baseline-shifted data. Results: A stratified 5-fold cross-validation using the Extra Trees algorithm successfully discriminated between PCR-confirmed Candidatus Liberibacter asiaticus-infected leaves and healthy controls, achieving an accuracy of 84.57% (95% confidence interval: 80.98%-88.15%). These results were obtained under field-like conditions and were further validated using headspace gas chromatography-mass spectrometry (HS-GC/MS), which revealed distinct VOC profiles for each group. Conclusion: This study demonstrates the potential of the electronic nose (e-nose) as a rapid, in-field screening tool capable of prioritizing samples for laboratory confirmation, thereby supporting effective HLB management.

Why it matches plant phenotyping methods柑橘葉の感染状態をVOCセンサーで直接推定する電子鼻を開発・評価しており、植物病害状態の取得方法が研究の中心である。

abstractThis study aims to develop and evaluate a compact electronic nose (e-nose) system equipped with metal-oxide semiconductor (MOS) sensors for the early detection of Candidatus Liberibacter asiaticus infection in citrus leaves through Volatile Organic Compound (VOC) analysis under field-like conditions.

The Accuracy and Efficiency of YOLO Algorithms in Identifying Plant Leaf Diseases →

Plant diseases are one of the biggest challenges of world agriculture, leading to enormous output losses and economic damages. Early and accurate detection of these diseases may help to increase crop yield, improve resource efficiency, decrease costs and environmental impact and assist the production of high-quality food. In recent years, deep learning (especially computer vision approaches) has become a strong tool for a number of tasks such as picture classification, segmentation and object detection. Such techniques include the You Only Look Once (YOLO) family of neural networks, a state-of-the-art technology for accurate object detection. In this work, we use YOLOv5, YOLOv7 and YOLOv8 models for citrus disease detection with the CCL’20 dataset. During training, a number of data augmentation techniques are used to improve the model performance, such as picture translation, scaling, flipping and mosaic augmentation. The model performance was evaluated using the Mean Average Precision (mAP) for Intersection over Union thresholds from 50% to 95% (mAP@50–95). The results showed that the YOLOv8 model performed better than the other variations, with significant improvements compared to the benchmarks reported in previous studies. After hyper-parameter adjustment, the improved model reached a mAP@50-95 of 96.1% on the test set for detection of the citrus diseases. The model attained the mAP@50-95 of 95.3%, 96.0% and 97.0% for Anthracnose, Melanose and Bacterial Brown Spot respectively for each disease. Furthermore, the model could reliably identify both single and many cases of the same and different diseases inside a single image, illustrating the robustness of recent YOLO architectures. Finally, the trained YOLOv8 model has been successfully installed into the Roboflow platform which is ready for practical applications in citrus disease monitoring.

Why it matches plant phenotyping methods柑橘葉の病害状態を画像から検出・分類するYOLO手法の開発と性能比較が中心であり、植物病害フェノタイピング手法に該当する。

abstractIn this work, we use YOLOv5, YOLOv7 and YOLOv8 models for citrus disease detection with the CCL’20 dataset.

LDTC-YOLO: A Lightweight Detection Model for Typical Citrus Leaf and Fruit Diseases in Real Orchard Environments →

Accurate detection of citrus leaf and fruit diseases is important for precision orchard management. However, real orchard images often contain small disease symptoms, leaf and fruit overlap, illumination variation, and cluttered backgrounds, making reliable detection challenging. This study proposes LDTC-YOLO, a lightweight YOLOv8n-based detection model for typical citrus leaf and fruit diseases in real orchard environments. To improve detection accuracy and model compactness, LDTC-YOLO integrates an Adaptive Feature Pyramid Network (AFPN) for cross-level feature fusion, Coordinate Attention (CA) for disease-region feature enhancement, a Lightweight Shared Convolutional Detection (LSCD) head for reducing parameter redundancy, and Wise-IoU (WIoU) for bounding-box regression optimization. In addition, a self-collected handheld citrus disease dataset, HOCD-4, was constructed using close-range smartphone images captured in real orchards. The dataset covers leaf and fruit symptoms of four typical citrus diseases: Huanglongbing/citrus greening (HLB), black spot, canker, and melanose. Experimental results show that LDTC-YOLO achieved precision, recall, mAP@0.5, and mAP@0.5:0.95 values of 0.915, 0.843, 0.894, and 0.648, respectively. Compared with YOLOv8n, LDTC-YOLO reduced parameters, GFLOPs, and model size from 3.006 M to 1.887 M, 8.1 to 7.4, and 5.97 MB to 3.83 MB, while increasing inference speed from 43.14 FPS to 47.45 FPS. These results indicate that LDTC-YOLO improves detection performance while maintaining a compact and efficient model profile, providing a potential reference for citrus disease detection under real orchard imaging conditions.

Why it matches plant phenotyping methods柑橘葉・果実の病徴を画像から検出する軽量モデルと実圃場データセットを開発・評価しており、植物の病害状態を推定するフェノタイピング手法が中心である。

abstractThis study proposes LDTC-YOLO, a lightweight YOLOv8n-based detection model for typical citrus leaf and fruit diseases in real orchard environments.

Hybrid IGWO-Dingo optimized DeMoHybridNet model for multi-class leaf disease identification. →

To achieve efficient crop management, exact plant disease detection in leaves is required. This study proposes DeMoHybridNet (Bottleneck Reduction Fusion) for the automated classification of Corn, Apple, Citrus, and Mango crop diseases. Input images are processed through augmentation and resizing, and then features are learned using DenseNet-201 and MobileNetV2. Global Average Pooling is applied, which produces condensed features. The features are then compressed using bottleneck layers of 512 features. The features are concatenated and classified by Random Forest (RF) classifier. To further improve the performance, a hybrid meta-heuristic method called IGWO-DOA (Improved Grey Wolf Optimization-Dingo Optimization Algorithm) is used to optimize the hyperparameters of the model for better convergence and generalization. The proposed optimized model gives the classification accuracy is 98.56% for Corn, 98.99% for Apple, 97.83% for Citrus and 99.35% for Mango leaf dataset. Statistical analysis confirms its robustness and reliability, demonstrating its effectiveness for precision agriculture applications.

Why it matches plant phenotyping methods葉画像から植物の病害状態を自動分類する深層学習・特徴抽出・分類ワークフローが研究の中心であり、植物病害表現型の画像ベース推定手法に該当する。

abstractThis study proposes DeMoHybridNet (Bottleneck Reduction Fusion) for the automated classification of Corn, Apple, Citrus, and Mango crop diseases.

Precision and accuracy of tree height estimation in citrus orchards: a systematic investigation of manual, airborne LiDAR, SLAM LiDAR, AI-driven photogrammetry →

Abstract has not been obtained from indexed metadata or an accessible article page.

Why it matches plant phenotyping methods柑橘樹の樹高という植物形質の推定精度・正確度を、手測定、LiDAR、SLAM、AIフォトグラメトリで体系的に比較検証する研究であり、フェノタイピング手法の技術評価が中心です。

titlePrecision and accuracy of tree height estimation in citrus orchards: a systematic investigation of manual, airborne LiDAR, SLAM LiDAR, AI-driven photogrammetry

Interpretable Color–Texture–Shape Feature Fusion for RGB-Based Citrus Canker and Melanose Classification on Orange Fruit →

Abstract Citrus canker and melanose substantially reduce the visual quality and commercial value of orange fruit, yet routine diagnosis remains largely dependent on subjective visual inspection. This study presents an interpretable color–texture–shape feature fusion framework for RGB-based classification of healthy, canker-infected, and melanose-affected orange fruit. Each image was represented by a compact descriptor integrating HSV color histograms, Local Binary Pattern micro-texture features, and Histogram of Oriented Gradients edge–shape information, followed by normalized feature concatenation and one-vs-rest Logistic Regression classification. On a balanced held-out test set, the proposed pipeline achieved 93.93% overall accuracy and a macro-F1 score of approximately 0.94, with class-wise F1-scores of 0.927 for citrus canker, 0.933 for healthy fruit, and 0.958 for melanose. Error analysis showed that residual misclassifications were concentrated mainly along the canker–healthy boundary. These findings demonstrate that well-designed handcrafted descriptors can provide accurate, transparent, and diagnostically meaningful citrus fruit disease recognition.

Why it matches plant phenotyping methodsRGB画像から果実の色・テクスチャ・形状特徴を抽出し、病徴状態を分類する方法が研究の中心であるため、植物病害表現型の画像ベース手法として含める。

abstractThis study presents an interpretable color–texture–shape feature fusion framework for RGB-based classification of healthy, canker-infected, and melanose-affected orange fruit.

Field-Reliability Analysis of Lightweight Orange Disease Screening: Image-Quality Effects, Failure Taxonomy, and Practical Triage for Edge Deployment →

Abstract Reliable edge-based citrus disease screening requires more than high benchmark accuracy; it must identify when an image and its prediction are trustworthy under field conditions. This study analyzed a balanced three-class orange image corpus comprising healthy fruit, citrus canker, and melanose using a lightweight HSV–LBP–HOG representation and calibrated confidence profiling. Rather than treating all classifications as equally actionable, the analysis linked prediction confidence with image-quality limitations and residual error patterns. The results showed that difficult cases were concentrated around the early canker–healthy boundary, where weak chromatic changes, subtle rind texture, shadowing, blur, and non-disease surface defects reduced diagnostic reliability. A quality-aware triage scheme was therefore established to separate direct acceptance, image recapture, and expert or local refinement. The findings support a practical field-reliability framework for low-cost orange disease screening on mobile and edge devices.

Why it matches plant phenotyping methods柑橘病害の画像分類について、画像品質、失敗分類、信頼度校正、現場トリアージを中心に評価しており、植物の病徴・病害状態を推定する手法の検証が主題である。

abstractThis study analyzed a balanced three-class orange image corpus comprising healthy fruit, citrus canker, and melanose using a lightweight HSV–LBP–HOG representation and calibrated confidence profiling.

Edge-Ready Citrus Disease Screening Using Classical Vision Features: Computational Efficiency, Robustness, and Threshold-Tunable Canker Detection →

Abstract Edge-based citrus disease screening requires not only accurate recognition but also low latency, modest memory use, robustness to imperfect image acquisition, and flexible decision thresholds for practical field operation. This study evaluated a lightweight classical vision pipeline as an operational screening engine for orange fruit disease detection under CPU-only deployment constraints. RGB images were represented using HSV color histograms, Local Binary Pattern texture descriptors, and Histogram of Oriented Gradients shape features, followed by one-vs-rest Logistic Regression classification. Beyond classification accuracy, the system was assessed through computational profiling, perturbation robustness, descriptor-level accuracy–latency trade-offs, and threshold-tunable canker detection. The pipeline achieved macro-F1 ≈ 0.94 while requiring only ~ 0.482 ms/image for feature extraction, negligible classification latency, ~ 106 MB RAM, and ~ 6% CPU utilization. Robustness analysis showed stable performance under brightness, contrast, crop, rotation, and moderate noise perturbations. These findings support classical feature-based vision as a practical, transparent, and resource-efficient edge-screening strategy for sustainable citrus disease monitoring.

Why it matches plant phenotyping methods柑橘果実の病徴・かんきつかんきつ類かいよう病を画像から検出する古典的コンピュータビジョン手法が中心で、精度、頑健性、計算性能、閾値調整を評価しているため。

abstractThis study evaluated a lightweight classical vision pipeline as an operational screening engine for orange fruit disease detection under CPU-only deployment constraints.

Robotic Tactile Sensing for Early Detection of Frost-Damaged Citrus Fruits with Pressure-Vibration Multimodal Fusion. →

Early-stage frost damage in citrus fruits is difficult to detect because external symptoms are often weak or absent, hindering intelligent robotic sorting in postharvest scenarios. To address this challenge, this study proposes a robotic multimodal tactile sensing approach inspired by human mechanoreception for frost-damage detection during grasping. A robotic gripper equipped with a 6×6 pressure matrix sensor and a piezoelectric vibration sensor was used to capture complementary tactile cues during standardized fruit handling, enabling the perception of subtle mechanical changes associated with early frost injury. Using 240 Citrus reticulata 'Hong Mei Ren' fruits under controlled experimental conditions, a Transformer-based multimodal fusion network was developed to jointly model pressure and vibration sequences for binary classification of normal and frost-damaged fruits. Across repeated stratified random-split experiments, the proposed method achieved a mean classification accuracy of 93.1%. Comparative experiments showed that the fusion model outperformed representative sequence-learning baselines, and ablation analysis confirmed that pressure-vibration fusion was more effective than either single modality alone. Attention-based temporal attribution further revealed that the most informative cues were concentrated in the initial contact and early loading stages, indicating the importance of early transient mechanical responses for frost-damage discrimination. Overall, the proposed approach demonstrates the feasibility of grasp-based robotic frost-damage detection under controlled experimental conditions.

Why it matches plant phenotyping methods柑橘果実の凍害状態を圧力・振動センサーで取得し、マルチモーダル融合により分類する手法の開発が中心であり、単なる生物学的実験の測定ではない。

abstractthis study proposes a robotic multimodal tactile sensing approach inspired by human mechanoreception for frost-damage detection during grasping.

Machine Learning-Based Non-Destructive Prediction of Juice Sac Granulation in Guanxi Honey Pomelo. →

The juice sac granulation of citrus fruits is a biological disorder that commonly occurs during the stages of growth, mature, and post-harvest, which severely affects the quality and reduces consumer acceptance of fruits. To explore the correlation between granulation and both external morphological characteristics and internal quality characteristics, 11 external and internal quality characteristics of Guanxi honey pomelo were collected and systematically analyzed by principal component analysis and linear regression. Then seven external quality characteristics and one critical characteristics, GR% were applied in machine learning modeling. The results indicated that several characteristics such as single fruit weight, single fruit volume, longitudinal diameter, and transverse diameter showed positive correlations with juice sac granulation rate (GR%), and were subsequently incorporated into classification model development. Among the five models evaluated, support vector machine demonstrated superior performance with a precision and recall rate of 100.00% and 100.00%, respectively, verifying its favorable accuracy and robustness. This research combined traditional statistical approaches with modern computational techniques, offering a reliable screening solution for juice sac granulation degree of Guanxi honey pomelo, which provided potential applicability in citrus processing industries and a theoretical foundation for non-destructive quality assessment.

Why it matches plant phenotyping methods果実の外観・内部特性から果肉粒化率を非破壊予測する機械学習モデルを開発・評価しており、植物状態の取得・推定手法が研究の中心です。

titleMachine Learning-Based Non-Destructive Prediction of Juice Sac Granulation in Guanxi Honey Pomelo.

A multi-scale parallel weighted fusion dynamic attention method for citrus leaf disease recognitions. →

To address the low detection accuracy caused by leaf occlusion, the loss of disease targets, and complex backgrounds in citrus leaf disease detection, this study proposes a leaf disease detection method termed DBG-DETR (a real-time detection transformer with DMGF, BDFF, and GSDT). Firstly, a DMGF-ResNet18 (dynamic multi-scale gating fusion block) is designed as the disease feature extraction module. By leveraging multiscale parallel depthwise separable convolutions, this module adaptively extracts and fuses rich disease-related features. Secondly, a GSDT (gated sparse dynamic transformer) is introduced to focus on deep features. Through a dynamic gating mechanism and Top-K sparse attention, GSDT reduces model parameters while enabling the network to concentrate on disease regions. Finally, a BDFF (bi-directional dense feature fusion module) is proposed to facilitate effective interaction between shallow and deep features, achieving efficient disease feature fusion. Experimental results on a real or chard dataset demonstrate that, compared with the baseline model, DBG-DETR improves P, mAP mmAP, R and F1 by 3.31%, 3.40%, 4.11%, 3.89% and 3.59%, respectively, while reducing the number of parameters by 3.78 MB. These results indicate that the proposed method significantly enhances disease detection performance in complex background environments and provides reliable technical support for intelligent citrus orchard management.

Why it matches plant phenotyping methods柑橘葉の病害領域を画像から検出する手法を提案・実験評価しており、植物病害状態の取得・推定が研究の中心である。

abstractthis study proposes a leaf disease detection method termed DBG-DETR

Deep learning-based citrus plant disease classification using a computationally efficient CNN model. →

Recent advancements in domain-specific classification methods have demonstrated the remarkable performance of deep learning in comparison to traditional machine learning techniques. This study develops a computationally efficient Convolutional Neural Network (CNN) model tailored for citrus plant disease classification, achieving performance comparable to that of the pretrained InceptionV3 model. A custom five-layer CNN model is constructed to classify citrus plant diseases into healthy and diseased categories using images collected from citrus orchards in Northen India. The model has been further validated using images sourced from GitHub and the Kaggle database. The proposed method surpasses classical machine learning approaches in accuracy and computational efficiency, achieving classification accuracies of 92.59%. The training time of the proposed CNN AgriVision-L5 is reduced by 50%, respectively, compared to the InceptionV3 model, demonstrating their computational efficiency. The proposed methodology offers significant advancements in plant disease management and sustainable agriculture, aligning with Sustainable Development Goals like SDG2, SDG9, and SDG12.

Why it matches plant phenotyping methods柑橘病害を画像から分類するCNNを開発し、外部画像で検証しており、植物の病害状態を推定する方法が研究の中心です。

abstractThis study develops a computationally efficient Convolutional Neural Network (CNN) model tailored for citrus plant disease classification

Deep learning based instance segmentation of mandarin fruit slices for precision assessment and morphological quantification. →

Accurate instance segmentation of mandarin fruit slices is essential for quantifying segment morphology and central core structure, which are key traits in cultivar evaluation, fruit quality assessment, and postharvest application. Manual measurement of these anatomical features, however, is time-consuming and prone to inconsistency. In this study, we developed an automated segmentation framework based on YOLOv8 to detect and semantic segment the fruit segments and the central core in high-resolution mandarin transversely cut images. The model was trained on a curated datasetderived from 58 original high-resolution cross-sectional images (5100 × 7019 pixels), which were systematically partitioned into 280 cropped sub-images (1712 × 1778 pixels), each containing a single complete citrus slice, and demonstrated excellent performance. YOLOv8 achieved near-perfect detection metrics, with bounding box metrics precision ~ 0.997, recall ~ 1.00, mAP50 ~ 0.995, and mAP50-95 ~ 0.922. Semantic segmentation accuracy was similarly strong, with precision = 0.997, Recall ~ 1.00, mAP50 ~ 0.995, and mAP50-95 = 0.965. Training and validation losses converged steadily, indicating stable learning without overfitting. The nonsignificant differences between YOLOv8-predicted measurements and ground truth data, together with low mean absolute error (MAE) values, demonstrate that the model not only performs well in semantic segmentation metrics but also maintains high accuracy in quantitative measurements, which is critical for cultivar discrimination and genetic studies. Our work provides a robust, high-precision, and reproducible framework for mandarin fruit slice phenotyping, offering significant potential for applications in agricultural research, breeding programs, and automated fruit quality evaluation.

Why it matches plant phenotyping methodsマンダリン果実スライスの形態・中心部構造を画像分割で定量する手法を開発し、精度検証と実測値比較を行っており、植物フェノタイピング手法が中心である。

abstractwe developed an automated segmentation framework based on YOLOv8 to detect and semantic segment the fruit segments and the central core

Hybrid LSTM-edge correction architecture for physics-informed crop health monitoring in distributed agricultural robotics →

Agricultural robotics-enabled crop health monitoring faces critical trade-offs: standalone on-device models sacrifice accuracy for real-time responsiveness, while cloud-dependent approaches suffer from high latency and communication overhead. Additionally, data-driven models often lack biophysical plausibility, leading to unreliable predictions for agronomic decision-making under resource constraints. We propose a hybrid LSTM-edge correction architecture that hierarchically integrates lightweight Long Short-Term Memory (LSTM) networks on field robots with physics-informed neural networks (PINNs) at the edge. On-device LSTMs process localized sensor data (soil moisture, spectral reflectance) to generate initial crop stress probability estimates with minimal latency. Edge-based PINNs refine these predictions by embedding biophysical dynamics—modeled via coupled partial differential equations (PDEs) governing the soil-plant-atmosphere continuum (SPAC)—to ensure agronomic validity, mitigate sensor noise, and account for spatial variability. The framework is deployed on NVIDIA Jetson Nano (local inference) and AMD EPYC servers (edge processing), seamlessly integrating with existing farming infrastructures to replace rule-based thresholds with adaptive, physics-grounded control commands. A Fourier Neural Operator (FNO) optimizes the edge PINN’s computational efficiency for high-dimensional PDE solving. Experimental evaluations on two real-world datasets (soybean and citrus) demonstrate that the hybrid approach improves prediction accuracy by 18% compared to standalone LSTMs (F1-score: 0.89±0.02 for soybean, 0.83±0.03 for citrus) while maintaining real-time performance (end-to-end latency: 210 ms, energy consumption: 5.1 J/prediction). Field deployment on a 50-hectare soybean farm yields tangible agronomic benefits: 22% reduction in irrigation water usage, 18% fewer pesticide applications, and 95% system uptime under field conditions. The framework exhibits robust performance against sensor noise (≥80% accuracy at 30% noise-to-signal ratio) and outperforms cloud-based PINNs (72.8% lower energy consumption) and threshold-based methods (28–33% higher F1-score). This work advances distributed agricultural robotics by bridging data-driven machine learning and domain-specific physics, delivering a scalable, interpretable, and resource-efficient solution for precision agriculture. The hierarchical prediction-correction pipeline balances real-time responsiveness with biological plausibility, making it suitable for resource-constrained field robots. By integrating legacy sensors and adaptive actuation control, the architecture offers a practical pathway to upgrade existing farming systems, enabling data-informed interventions while reducing environmental impact.

Why it matches plant phenotyping methods作物ストレス状態を推定するLSTM・PINN・FNO統合パイプラインを開発し、実データで精度・遅延・ノイズ耐性を評価しており、植物状態の取得・推定手法が中心である。

abstractWe propose a hybrid LSTM-edge correction architecture that hierarchically integrates lightweight Long Short-Term Memory (LSTM) networks on field robots with physics-informed neural networks (PINNs) at the edge.

Systematic Investigation of Microstructural and Spectral Characteristics in Citrus Midrib for Huanglongbing Detection. →

Citrus Huanglongbing (HLB) disease is a devastating disease faced by the global citrus industry, for which there is currently no effective cure. This study systematically investigated the anatomical characteristics and infrared spectral properties of different microstructures (phloem, xylem, pith, and cortical tissues) in the midribs of healthy and HLB-infected citrus leaves. Scanning electron microscopy observations revealed obvious phloem breakage and massive starch granule accumulation in various tissues of HLB-infected samples. Using micro-Fourier transform infrared spectroscopy, the spectral acquisition parameters were optimized (slice thickness: 10 μm, spectral resolution: 8 cm -1 , spatial resolution: 10 μm × 10 μm, number of scans: 256), and in-situ spectral information from different tissues were obtained. The results showed significant changes in the intensity and position of absorption peaks in the fingerprint region (1,800 to 675 cm -1 ) of all tissues after HLB infection, particularly enhanced carbohydrate absorption at bands such as 1,099, 1,060, and 1,033 cm -1 , indicating that abnormal carbohydrate accumulation is a typical symptom of HLB. A principal component analysis score plot based on spectral data from the phloem demonstrates a clear spatial separation trend between healthy and HLB-infected samples, providing a theoretical basis and methodological support for the fast, early, nondestructive detection of citrus HLB disease.

Why it matches plant phenotyping methodsミクロFTIRの取得条件を最適化し、健全・HLB感染葉の組織スペクトルから病害状態を識別する手法を検討しており、植物病徴の非破壊フェノタイピングが中心である。

abstractUsing micro-Fourier transform infrared spectroscopy, the spectral acquisition parameters were optimized (slice thickness: 10 μm, spectral resolution: 8 cm -1 , spatial resolution: 10 μm × 10 μm, number of scans: 256), and in-situ spectral information from different tissues were obtained.

Dataset for orange fruit detection from UAV in citrus orchards. →

Accurate fruit detection in citrus orchards is essential for yield estimation, precision harvesting, and automated orchard monitoring. Although UAV-based imaging has become a powerful tool in precision agriculture, publicly available datasets for orange fruit detection remain scarce, particularly those integrating multispectral data under real field conditions. This lack of open resources limits the development and benchmarking of robust deep-learning models for cross-spectral and illumination-invariant detection. We present CampanetaOrangeFruit, a dataset acquired with a DJI Mavic 3 Multispectral UAV flying at 14 m above ground level over a commercial citrus orchard in Corbera, Valencia, Spain. The dataset comprises 550 synchronized captures (RGB + four multispectral bands: R, G, RE, NIR) for a total of 2750 images and 301,232 annotated orange instances. Each image includes YOLOv5-format annotations generated through a homography-based reprojection process, ensuring geometric consistency across spectral modalities. CampanetaOrangeFruit uniquely provides pixel-aligned, cross-spectral UAV imagery with fine-grained fruit-level annotations, enabling research on fruit detection, yield estimation, and domain adaptation in real-world orchard environments. It represents a valuable benchmark for advancing deep-learning approaches in precision agriculture and sustainable citrus production.

Why it matches plant phenotyping methods柑橘果実を対象としたUAV画像データセットとアノテーションを提供し、果実検出・収量推定モデルの開発およびベンチマークを中心課題とするため、植物フェノタイピング用データセットとして採用する。

abstractpublicly available datasets for orange fruit detection remain scarce

Simulation of citrus foliar gas exchange across diverse meteorological conditions: application of the optimal stomatal regulation method. →

Introduction The optimal stomatal regulation theory provides an eco-evolutionary framework for interpreting the trade-off between CO 2 uptake and water loss. This theory postulates that the marginal water cost of carbon gain ( λ=∂E/∂A ) remains approximately constant over short timescales, thereby offering a mechanistic basis for predicting stomatal behavior and gas exchange. Methods In this study, leaf-level meteorological variables and gas exchange parameters of orchard citrus were measured throughout the entire phenological period during 2021-2022. We developed a family of optimal stomatal conductance-based models (OSCMs), comprising six forms: Rubisco-limited forms (OSCvc and OSCvcd), RuBP-regeneration-limited forms (OSCvj and OSCvjd), and combined forms that dynamically select the prevailing biochemical limitation (OSC and OSCd). Results The key parameter λ was estimated daily and averaged over the entire phenological period. Using daily λ inputs, the three models produced stomatal conductance ( g s ) with accuracies ranked as OSCvjd (R 2 = 0.73) > OSCd (0.63) > OSCvcd (0.40). When a long-term constant λ was applied, model performance declined with accuracies ranked as OSCvj (0.66) > OSC (0.52) > OSCvc (0.38). Discussion The OSC model also produced intercellular CO 2 concentration ( c i ) and photosynthesis ( A ) reasonably well (R 2 = 0.78 and 0.48, respectively). Under moderate meteorological conditions (air temperature 30-40 °C and vapor pressure deficit 1-2 kPa), the OSC model showed its best performance with a mean absolute relative error of 35.2% for g s estimation. Overall, the OSCMs provided a mechanistic approach to simulate citrus leaf gas exchange requiring minimal species-specific traits and routine meteorological inputs. This modeling strategy supports rapid assessment of plant physiological status and estimation of foliar carbon-water fluxes in orchard management under subtropical climates.

Why it matches plant phenotyping methods柑橘葉のガス交換・気孔コンダクタンスを推定するモデル群を開発し、実測値との精度比較で検証しており、植物生理形質の取得・推定法が研究の中心である。

abstractWe developed a family of optimal stomatal conductance-based models (OSCMs), comprising six forms

A hybrid convolution and attention-based framework with visual explanation for fruit disease identification. →

The objective of this study is to create a highly accurate and interpretable deep learning (DL) model for the multi-class classification of fruit using convolutional and transformer architectures. The classification performance can be enhanced by making sure that the used technique is explainable and interpretable. This research data was obtained from Kaggle which contains images of banana, grape, lemon, mango, and strawberry fruit classes. The total data was divided into 70:15:15 for training, validating and testing. To ensure consistent size and quality, all images were pre-processed before use. This study considered four pretrained models namely RegNetY-B3-GE, DarkNet53-SCSE, BEiT, and PVTv2 for performance assessment. We proposed a lightweight hybrid (convolution plus attention-based) CoAT-AgriLite model for fruit disease classification which extracts local lesion features and global context. Transferring training and data augmentation technique was utilized during training for better performance. To ensure interpretability of model decisions, Gradient-weighted Class Activation Mapping (Grad-CAM) which captures the discriminative regions from the input images for model predictions. Among all evaluated models, the proposed model achieved the highest classification accuracy of 99.37% on the testing dataset. Comparative results demonstrated that the proposed model outperformed other pretrained models in terms of precision, recall, and F1-score, confirming its robustness and effectiveness in real-world agricultural classification tasks. The experimental findings validate that the proposed model not only achieves superior classification accuracy but also provides interpretability through Grad-CAM visualizations. This hybrid framework offers a promising solution for intelligent and transparent fruit classification systems, with potential applications in precision agriculture and automated sorting systems.

Why it matches plant phenotyping methods果実病害を画像から分類する深層学習フレームワークを開発・比較し、病斑特徴の抽出とGrad-CAMによる説明性を評価しており、植物の病害状態の画像計測が中心である。

titleA hybrid convolution and attention-based framework with visual explanation for fruit disease identification.

A high-performance biocompatible biomass-based fish gelatin organohydrogel strain sensor for long-term accurate plant growth monitoring →

Organohydrogel-based plant strain sensors hold significant potential for enabling accurate and real-time monitoring of plant growth processes. However, existing strain sensors typically face challenges such as inferior biocompatibility, trade-off between sensing performance and mechanical properties, as well as poor long-term stability, leading to inaccurate monitoring and plant tissue damage and thus hindering their practical applications. Herein, we propose a synergistic metal ion and multiple hydrogen bond dual crosslinking strategy to develop a biomass-based fish gelatin organohydrogel as a strain sensing material. The resultant organohydrogel simultaneously exhibits excellent mechanical properties (Young's modulus of 99.9 kPa and strong adhesiveness of 60 kPa), high sensing performance (GF = 2.13, stable response across a wide temperature range from -80 °C to 25 °C), outstanding plant tissue and human cell biocompatibility, and long-term stability (over 5000 loading-unloading cycles under 100% strain), demonstrating superior overall performance to most existing organohydrogels. To harness these unique material performances, we fabricate a sandwich-structured plant strain sensor for long-term monitoring of plant growth. The fabricated strain sensor enables successful real-time monitoring of the growth dynamics of lotus stems and pomelo fruits with high accuracy and long-term stability up to three weeks. Our novel design strategy of high-performance organohydrogels enables high-fidelity plant growth monitoring, unlocking new potentials for advancing data-driven smart and precision farming practices.

Why it matches plant phenotyping methods植物成長を長期・リアルタイムに測定するひずみセンサーの材料設計、性能評価、植物での検証が研究の中心であり、植物フェノタイピング手法に該当する。

abstractwe propose a synergistic metal ion and multiple hydrogen bond dual crosslinking strategy to develop a biomass-based fish gelatin organohydrogel as a strain sensing material

Fusion of multimodal features acquired by custom-developed computer vision and electronic nose equipment for the detection of Huanglongbing at various symptomatic stages →

Citrus is widely loved for its rich nutritional value and unique flavor, and also occupies an important place in agriculture and the economy. The citrus industry has suffered severe losses in recent years due to the proliferation of citrus Huanglongbing (HLB). The transmission of HLB occurs via the Asian citrus psyllid insect vector and grafting practices, with no efficacious therapeutic intervention identified to date, aside from mitigating its dissemination through prompt identification and eradication of infected citrus trees. Detection of HLB is difficult due to its incubation period and the variety of symptoms at different stages of infection. Hence, there exists a pressing requirement for a detection methodology capable of integrating multi-level features of HLB to facilitate precise identification of the disease across various stages of infection. Most of the existing assays use single sensing, which leads to limitations and incompleteness in identifying specific markers induced by HLB. In this study, the effective configuration and complementarity of multimodal sensory information is achieved by establishing a fusion and complementary mechanism at the level of pre-processing and analyzing multisource information. The extraction of computer vision and electronic nose features of citrus leaves was realized using custom-developed portable detection devices. The performance of HLB detection was compared on different datasets obtained by multimodal feature fusion methods which include direct fusion method, stepwise fusion method and the improved Recursive Feature Elimination and Cross Validation (RFECV) feature selection method. The improved RFECV feature selection method uses the RFECV algorithm for each classification step in the delineated stepwise classification model and performs the feature set preference by cross-validation. The final improved RFECV feature selection method worked best for fusion of visual and olfactory features with an accuracy of 95.38% for HLB samples at various symptomatic stages, with 94.23% for early stage HLB and 94.12% for Zn Def. & HLB-positive samples. Multimodal feature fusion for feature acquisition proved to be superior to feature acquisition from a single sensing source, with enhanced fusion of HLB-induced feature sets at the visual and olfactory levels. It helps to improve the stability of the HLB measurement model to achieve the detection of HLB samples in complex environments. This method can provide generalized technical support for the application of multi-source sensing information and multimodal feature fusion methods in plant disease detection.

Why it matches plant phenotyping methods柑橘葉の症状を対象に、カスタム開発した画像処理・電子鼻装置とマルチモーダル特徴融合によるHLB検出法を開発・評価しており、植物病態の取得・推定が研究の中心である。

abstractThe extraction of computer vision and electronic nose features of citrus leaves was realized using custom-developed portable detection devices.

A spectral-physiological feature fusion model for the early detection of anthracnose in citrus leaves →

Citrus anthracnose is a destructive fungal disease caused by Colletotrichum gloeosporioides, which causes leaf damage, fruit rot, and yield loss in citrus production. This study proposes an early detection method for citrus leaf anthracnose that integrates spectral and physiological data. Artificial inoculation experiments showed that the infected leaves exhibited yellowish-brown lesions, and the reflectance derived from visible-near-infrared (VNIR) spectroscopy and Fourier transform near-infrared (FTNIR) spectroscopy significantly decreased. Stomatal conductance and photosynthetic rate declined 4 days after inoculation. Physiological damage to leaves caused by fungal infection was more severe than mechanical damage. Three wavelength extraction algorithms [particle swarm optimization (PSO), bootstrapping soft shrinkage (BOSS), and least absolute shrinkage and selection operator (LASSO)] were combined with three machine learning models [artificial neural network (ANN), k-nearest neighbor (KNN), and categorical boosting (CatBoost)] to perform feature-level fusion on spectral data, photosynthetic parameters, and vegetation indices to improve classification accuracy. The fusion model had high classification accuracy (0.958–0.989) and Matthews correlation coefficient (MCC) (0.917–0.978). The model achieved the best performance in distinguishing leaves with early disease symptoms from healthy leaves, with an accuracy of 0.989, an F1 score of 0.989, and an MCC of 0.978. This research provides a reliable theoretical basis and technical support for the precise identification and early prevention and control of citrus anthracnose.

Why it matches plant phenotyping methods柑橘葉の病害状態をスペクトル・生理計測から推定する早期検出法を開発し、特徴抽出と機械学習モデルの性能を評価しており、植物表現型取得・推定が中心である。

abstractThis study proposes an early detection method for citrus leaf anthracnose that integrates spectral and physiological data.

Primary metabolomics analyses and detection of citrus “huanglongbing” disease based on UHPLC-MS/MS and machine learning →

‘Candidatus Liberibacter asiaticus’ is the major agent associated with citrus “huanglongbing” (HLB) disease, which is the most destructive citrus disease and has caused serious losses to citrus industry worldwide. Ultra-high performance liquid chromatography-tandem mass spectrometry (UHPLC-MS/MS)-based nontargeted metabolomics and machine learning algorithms were developed for identifying HLB disease in different citrus varieties and growing seasons. In this study, 52 (28 up-regulated and 24 down-regulated) and 33 (26 up-regulated and 7 down-regulated) differential metabolites were screened in Navel orange (Citrus sinensis Osbeck) and Ponkan (Citrus reticulata Blanco cv. Ponkan) leaves, respectively. The variable importance in projection (VIP) algorithm was then used to select the common differential metabolites in HLB diseased samples, and a total of 19 differential metabolite variables were obtained from Navel orange and Ponkan varieties (mainly including primary metabolites such as D-ribose, D-threonate, L-ornithine). Finally, support vector machine (SVM) model based on the metabolites with significant features performed the best for the prediction of citrus HLB disease, with a classification accuracy of 100 %. The results showed that the proposed method was able to provide important and common information about citrus host-'Ca. L. asiaticus' interactions. They also demonstrated that combing untargeted metabolomics with machine learning can be effective tools for distinguishing citrus HLB infection (from asymptomatic to symptomatic) in different growing stages and cultivars.

Why it matches plant phenotyping methodsUHPLC-MS/MSメタボロミクスと機械学習を組み合わせ、柑橘のHLB感染状態を識別する手法を開発・評価しており、病害状態の推定が中心的な貢献である。

abstractUltra-high performance liquid chromatography-tandem mass spectrometry (UHPLC-MS/MS)-based nontargeted metabolomics and machine learning algorithms were developed for identifying HLB disease in different citrus varieties and growing seasons.

Utilizing deep learning models for early detection and classification of fruit diseases: towards sustainable agriculture and enhanced food quality. →

Productivity and quality of food are crucial for populations around the world. However, food faces challenges due to the threats of fruit diseases, which lead to poor food quality. Therefore, early detection and classification of fruit diseases are important to help farmers detect and overcome these diseases, thereby improving food quality and productivity. One of the biggest challenges in the agriculture field is classifying and detecting fruit diseases using traditional manual visual grading. As a result, deep learning and computer vision models have emerged as new methods for visual grading, offering higher accuracy in classification and detection. This study proposes deep learning models for fruit disease detection and classification in the early stages. Five deep learning models are used: Convolutional Neural Network (CNN), DenseNet121, EfficientNetB3, Xception, and ResNet50. These models are applied to detect six types of fruit diseases, including orange, grape, mango, guava, apple, and banana plant diseases. Image preprocessing and data augmentation techniques were employed for image processing. The results show accuracies of 96.25%, 99.14%, 96.17%, 94.06%, 96.72%, and 99.33% for the CNN, EfficientNetB3, ResNet50, DenseNet121, ResNet50, and EfficientNetB3 models, respectively, for detecting orange, grape, mango, banana, guava, and apple plant diseases. We compared our models with other deep learning models, and the model that utilized image preprocessing and data augmentation techniques demonstrated higher accuracy and performance. We recommend the EfficientNetB3 model for fruit disease detection based on these results.

Why it matches plant phenotyping methods果実植物の病害状態を画像から検出・分類する深層学習手法の開発と比較が研究の中心であり、植物病害フェノタイピングに該当する。

abstractThis study proposes deep learning models for fruit disease detection and classification in the early stages.

Rapid and accurate fruit volume estimation using a top-mounted mirror-based imaging system →

Abstract Size is one of the important external quality criteria of fruit and is often used in grading and sorting processes. Computer vision offers an effective solution for quickly and non-destructively estimating fruit volume. Several approaches involving three-dimensional (3D) reconstruction models of irregularly shaped fruit from multiple side-view images have been employed to enhance estimation accuracy. However, these approaches tend to be expensive and computationally complex. This study proposes a low-cost multi-view system for the rapid and accurate estimation of fruit volume. Using a top-mounted mirror-based setup, the system captures multiple fruit surfaces with a single camera, eliminating the need for rotation or multiple cameras. The captured multi-view image is analyzed using a global thresholding technique to calculate the multi-view area. A simple linear regression model is then built to estimate the fruit’s volume based on multi-view area, without requiring 3D model reconstruction. The proposed system was successfully tested for estimating the volume of two irregularly shaped fruits (pomelo and mango), achieving high coefficients of determination (0.986 and 0.988, respectively). Additionally, the system was tested with different fruit orientations, and the results showed that orientation did not affect volume estimation. These results demonstrate that this approach has strong potential not only for fruit quality assessment but also for other irregularly shaped solid objects where rapid and accurate volume estimation is needed.

Why it matches plant phenotyping methods果実の体積という植物器官形質を、鏡面マルチビュー画像と画像解析・回帰で推定する手法を開発・検証しており、形質取得法が研究の中心である。

abstractThis study proposes a low-cost multi-view system for the rapid and accurate estimation of fruit volume.

Deep Learning based Orange Crop Disease Detection Using Image based Intelligent Framework for Precision Monitoring →

Orange crop production is highly vulnerable to fungal, bacterial, and nutrient-related diseases that significantly reduce yield quality and economic productivity. Traditional manual inspection methods are time-consuming, subjective, and often ineffective for early disease diagnosis, creating a need for automated and intelligent monitoring solutions. This study proposes a deep learning–based image-driven framework designed to accurately detect major orange crop diseases using high-resolution leaf and fruit images captured in real field conditions. The methodology integrates image enhancement, segmentation using K-means clustering and Canny edge detection, and preprocessing steps such as resizing, normalization, augmentation, and class balancing. A curated dataset of 3,000 images across six classes—including canker, greening, melanose, black spot, nutrient deficiency, and healthy samples—was used to train multiple CNN architectures (AlexNet, VGG19, and Xception) and a fuzzy rank-based ensemble model. Experimental results demonstrate that the proposed enhanced framework outperforms conventional methods, achieving 96.51% accuracy with the ensemble model, while individual models such as Xception and VGG19 achieve 92.25% and 90.34% accuracy, respectively, confirming its effectiveness for precision disease monitoring in orange orchards.

Why it matches plant phenotyping methodsオレンジ葉・果実画像から植物の病徴・病害状態を推定する画像解析・深層学習フレームワークが研究の中心であり、病害フェノタイピング手法に該当する。

abstractThis study proposes a deep learning–based image-driven framework designed to accurately detect major orange crop diseases using high-resolution leaf and fruit images captured in real field conditions.

Detection of Citrus Plant Leaf Detection Using Non – Imaging Data →

Plant leaf diseases pose a significant challenge to agricultural productivity, necessitating efficient and accurate detection methods. This study presents an integrated approach combining deep learning (CNN) and machine learning (SVM, Random Forest, KNN, Logistic Regression) for plant disease classification using imaging and spectral data. The proposed system processes leaf images to detect unhealthy regions, extracting statistical features such as contrast (10.59), entropy (4.31), and mean intensity (137.24) to assess disease severity. CNN-based models demonstrated strong training accuracy but exhibited overfitting in validation performance. Among machine learning models, SVM and Logistic Regression achieved the highest accuracy (70%), while Random Forest performed moderately (54%), and KNN struggled (39%) due to high-dimensional spectral complexities. Confusion matrices revealed that Healthy and Greening categories often overlapped, leading to misclassifications. The findings suggest that a hybrid deep learning + machine learning approach enhances classification accuracy by leveraging both image-based and spectral features. Future improvements involve ensemble learning, better feature engineering, and real-time field deployment for automated disease detection. This research provides a scalable and effective solution for precision agriculture, enabling early disease diagnosis and improved crop health monitoring.

Why it matches plant phenotyping methods画像およびスペクトルデータから葉の病徴・重症度を推定する分類手法を開発・評価しており、植物表現型の取得が中心的な貢献である。

abstractThis study presents an integrated approach combining deep learning (CNN) and machine learning (SVM, Random Forest, KNN, Logistic Regression) for plant disease classification using imaging and spectral data.

Estimation of Citrus Leaf Relative Water Content Using CWT Combined with Chlorophyll-Sensitive Bands. →

In citrus cultivation practice, regular monitoring of leaf leaf relative water content (RWC) can effectively guide water management, thereby improving fruit quality and yield. When applying hyperspectral technology to citrus leaf moisture monitoring, the precise quantification of RWC still needs to address issues such as data noise and algorithm adaptability. The noise interference and spectral aliasing in RWC sensitive bands lead to a decrease in the accuracy of moisture inversion in hyperspectral data, and the combined sensitive bands of chlorophyll (LCC) in citrus leaves can affect its estimation accuracy. In order to explore the optimal prediction model for RWC of citrus leaves and accurately control irrigation to improve citrus quality and yield, this study is based on 401-2400 nm spectral data and extracts noise robust features through continuous wavelet transform (CWT) multi-scale decomposition. A high-precision estimation model for citrus leaf RWC is established, and the potential of CWT in RWC quantitative inversion is systematically evaluated. This study is based on the multi-scale analysis characteristics of CWT to probe the time-frequency characteristic patterns associated with RWC and LCC in citrus leaf spectra. Pearson correlation analysis is used to evaluate the effectiveness of features at different decomposition scales, and the successive projections algorithm (SPA) is further used to eliminate band collinearity and extract the optimal sensitive band combination. Finally, based on the selected RWC and LCC-sensitive bands, a high-precision predictive model for citrus leaf RWC was established using partial least squares regression (PLSR). The results revealed that (1) CWT preprocessing markedly boosts the estimation accuracy of RWC and LCC relative to the original spectrum (max improvements: 6% and 3%), proving it enhances spectral sensitivity to these two indices in citrus leaves. (2) Combining CWT and SPA, the resulting predictive model showed higher inversion accuracy than the original spectra. (3) Integrating RWC Scale7 and LCC Scale5-2224/2308 features, the CWT-SPA fusion model showed optimal predictive performance (R 2 = 0.756, RMSE = 0.0214), confirming the value of multi-scale feature joint modeling. Overall, CWT-SPA coupled with LCC spectral traits can boost the spectral response signal of citrus leaf RWC, enhancing its prediction capability and stability.

Why it matches plant phenotyping methods柑橘葉の相対含水量という植物生理形質を、ハイパースペクトルデータ、CWT、SPA、PLSRで推定する方法の開発・評価が中心であり、ルーチン測定ではない。

abstractA high-precision estimation model for citrus leaf RWC is established, and the potential of CWT in RWC quantitative inversion is systematically evaluated.

Artificial intelligence‐powered plant phenomics: Progress, challenges, and opportunities →

Abstract Artificial intelligence (AI), a key driver of the Fourth Industrial Revolution, is being rapidly integrated into plant phenomics to automate sensing, accelerate data analysis, and support decision‐making in phenomic prediction and genomic selection. This perspective paper synthesizes current advances, identifies major barriers, and proposes future directions to realize the transformative potential of AI‐enabled plant phenomics. We first provide an overview of AI technologies with the potential to address key challenges in phenomics, from data collection to phenotypic trait extraction and environmental sensing. We then present three case studies focusing on specialty crops (blueberry [ Vaccinium corymbosum L.] mechanical harvestability traits, strawberry [ Fragaria × ananassa (Duchesne ex Weston)] production, and citrus [ Citrus L.] disease) to illustrate practical applications of AI‐driven phenomics. Moreover, we highlight future perspectives and opportunities for further research and innovation. These include large foundation models, real‐time inference on edge devices, explainable AI, generative AI and digital twins, AI‐enhanced multi‐omics, agentic AI, and knowledge‐guided and data‐driven hybrid approaches. Finally, we discuss key challenges and limitations of applying AI to plant phenomics, including data curation, model generalization and bias, and ethical considerations related to equitable access to AI tools.

Why it matches plant phenotyping methods植物フェノミクスにおけるAIセンシング・形質抽出を主題とする展望論文であり、方法論のレビューとして中心的に扱っている。

abstractWe first provide an overview of AI technologies with the potential to address key challenges in phenomics, from data collection to phenotypic trait extraction and environmental sensing.

ReDuCit project - poster →

Title: ReDuCit Project Poster: Building a More Sustainable Agriculture Based on Robust, Scalable Applications of Controlled Deficit Irrigation Strategies Description: This poster presents the ReDuCit project, a 24-month initiative focused on developing sustainable irrigation strategies for citrus crops in the Guadalquivir River Basin (Spain). The project addresses the critical challenge of maintaining agricultural productivity while reducing water consumption in a region where citrus represents 5.1% of the irrigated area but accounts for 9.7% of water demand, and where climate projections indicate a 10% reduction in water availability by 2039. Project objectives include: Designing and validating an integrated system for monitoring water status and production in citrus crops Establishing a replicable and scalable model applicable to other crops Developing a digital platform for optimized irrigation management Implementing a robust Regulated Deficit Irrigation Control model capable of reducing water consumption by 15-25% Technical approach: The project combines water status monitoring (using sap flow sensors, trunk stem dendrometers, and microtensiometers on reference trees), production tracking (through autonomous RGB-D cameras with AI for fruit detection and measurement), and an integrated digital platform providing real-time data collection, automated irrigation recommendations, and personalized management alerts. Expected impact: Potential water savings of 60 million m³/year in citrus crops from the Guadalquivir region alone, representing 20% of the required reduction in the agricultural sector by 2039. Consortium: OnTech Innovation, Rovimatica, Universidad de Sevilla, Soltel Group Funding: Co-financed by European Funds through Junta de Andalucía and the Spanish Ministry of Finance Validation: Real-world testing in collaboration with the Irrigation Community of the Lower Guadalquivir Valley Bilingual poster (English/Spanish) Keywords: precision agriculture, deficit irrigation, water management, citrus crops, digital agriculture, IoT sensors, artificial intelligence, sustainability, Guadalquivir, smart farming

Why it matches plant phenotyping methods柑橘の水分状態と果実の検出・計測を行うセンサー／RGB-D・AI統合システムの設計・検証がプロジェクトの中心であり、植物状態・果実形質の取得方法を含むため。

abstractDesigning and validating an integrated system for monitoring water status and production in citrus crops

ReDuCit project - poster →

Title: ReDuCit Project Poster: Building a More Sustainable Agriculture Based on Robust, Scalable Applications of Controlled Deficit Irrigation Strategies Description: This poster presents the ReDuCit project, a 24-month initiative focused on developing sustainable irrigation strategies for citrus crops in the Guadalquivir River Basin (Spain). The project addresses the critical challenge of maintaining agricultural productivity while reducing water consumption in a region where citrus represents 5.1% of the irrigated area but accounts for 9.7% of water demand, and where climate projections indicate a 10% reduction in water availability by 2039. Project objectives include: Designing and validating an integrated system for monitoring water status and production in citrus crops Establishing a replicable and scalable model applicable to other crops Developing a digital platform for optimized irrigation management Implementing a robust Regulated Deficit Irrigation Control model capable of reducing water consumption by 15-25% Technical approach: The project combines water status monitoring (using sap flow sensors, trunk stem dendrometers, and microtensiometers on reference trees), production tracking (through autonomous RGB-D cameras with AI for fruit detection and measurement), and an integrated digital platform providing real-time data collection, automated irrigation recommendations, and personalized management alerts. Expected impact: Potential water savings of 60 million m³/year in citrus crops from the Guadalquivir region alone, representing 20% of the required reduction in the agricultural sector by 2039. Consortium: OnTech Innovation, Rovimatica, Universidad de Sevilla, Soltel Group Funding: Co-financed by European Funds through Junta de Andalucía and the Spanish Ministry of Finance Validation: Real-world testing in collaboration with the Irrigation Community of the Lower Guadalquivir Valley Bilingual poster (English/Spanish) Keywords: precision agriculture, deficit irrigation, water management, citrus crops, digital agriculture, IoT sensors, artificial intelligence, sustainability, Guadalquivir, smart farming

Why it matches plant phenotyping methods灌漑管理プロジェクトだが、果実の検出・計測を行うRGB-Dカメラ／AIと、水分状態を測定するセンサーを統合したモニタリング基盤が技術的中核として明示されており、植物の生産・生理状態の表現型取得に該当する。

abstractDesigning and validating an integrated system for monitoring water status and production in citrus crops

Design and implementation of laser-light backscattering imaging system as a non-destructive technique for citrus taste evaluation →

Citrus fruit quality, particularly taste, plays a crucial role in consumer preference and marketability. Conventional taste tests, such as sensory panel assessments and chemical analysis, are time-consuming and destructive, underscoring the need for rapid and non-destructive evaluation methods. Therefore, this study aimed to design laser-light backscattering imaging (LLBI) system as a novel approach for evaluating citrus taste. A total of 150 Siamese citrus samples were collected from Cisurupan Orchards. Sensory evaluation was performed using Quantitative Descriptive Analysis by 20 trained panelists to classify citrus taste into two categories namely sour and sweet. Moreover, the LLBI system was developed using laser diodes at three wavelengths (450, 532, and 648 nm) to capture backscattering images. A ResNet50-based deep learning model was implemented to classify citrus samples, with the performance evaluated using accuracy and the area under the receiver operating characteristic curve (AUC). The results showed that the 648 nm wavelength yielded the highest classification performance, achieving accuracies of 98.968 % for training, 96.898 % for validation, and 96.759 % for testing. The corresponding AUC values were 0.9996, 0.9967, and 0.9961, respectively, confirming the model excellent predictive capability. LLBI demonstrates significant potential as a non-destructive, rapid, and objective technique for evaluating citrus sensory quality.

Why it matches plant phenotyping methods柑橘の味覚状態を非破壊画像から推定する撮像システムと深層学習手法の開発・評価が研究の中心であり、植物器官の品質状態を測定するフェノタイピング手法に該当する。

abstractTherefore, this study aimed to design laser-light backscattering imaging (LLBI) system as a novel approach for evaluating citrus taste.

Investigating the Association Between Citrus Huanglongbing and Chlorophyll Content Using Hyperspectral Detection. →

Huanglongbing (HLB) poses a severe threat to the sustainable citrus industry, causing significant alterations in the spectral reflectance and leaf chlorophyll content (LCC) of citrus leaves. This study investigates the quantitative relationship between spectral characteristics and LCC for the early detection of HLB in Mianju mandarin cultivars. We analyzed hyperspectral data from healthy and HLB-infected leaves, employing the least absolute shrinkage and selection operator (LASSO) method and spectral indices to select chlorophyll characteristic bands, and several machine learning models were used to estimate the LCC. The results indicate that: (1) HLB-infected leaves exhibit significantly different spectral reflectance, characterized by a distinct "blueshift of the red edge"; (2) a greater proportion of characteristic bands for HLB-infected leaves were located in the near-infrared region compared to healthy leaves; and (3) the LASSO-PLSR model demonstrated high predictive accuracy for LCC estimation-for healthy leaves (Rv 2 = 0.956, RMSEv = 0.675) and for HLB-infected leaves (Rv 2 = 0.816, RMSEv = 4.614)-with performance being notably superior for healthy leaves (Rv 2 difference of +0.146). This research establishes a systematic quantification between hyperspectral and chlorophyll content, suggesting that hyperspectral-based LCC estimation can serve as a reliable indirect indicator for the early detection of HLB, with substantial practical application potential.

Why it matches plant phenotyping methods柑橘葉のクロロフィル含量という植物形質をハイパースペクトルデータと機械学習で推定し、HLB早期検出への有効性を定量的に検証しており、フェノタイピング手法が中心です。

abstractemploying the least absolute shrinkage and selection operator (LASSO) method and spectral indices to select chlorophyll characteristic bands, and several machine learning models were used to estimate the LCC

PhenoCitrus: An automated platform to phenotyping morphological traits of citrus fruit →

Citrus breeding critically relies on precise phenotyping, yet existing RGB/3D phenotyping methods lack standardized workflows balancing affordability, accuracy, and efficiency. An integrated hardware–software pipeline is introduced to address this gap: (1) a custom imaging device for standardized top/side-view capture of the sample fruit, and (2) optimized computer vision algorithms using enhanced 3D Gaussian Splatting (3DGS) to extract 3D structural traits (surface area, volume) and 2D algorithms (e.g., Unet++ and YOLO variants) for 2D trait (width, length, oil cell number, peel/pulp color, peel thickness, and segment number) quantification. Our approach achieved average Pearson correlations above 0.9 between algorithmic and manual measurements across five key traits, confirming measurement precision. The resulting phenotypic profiles revealed biologically significant inter-trait correlations to inform trait-oriented selection decisions and the refinement of breeding strategies. Finally, we operationalized this workflow through user-friendly software, delivering an end-to-end solution that enables high-throughput, low-cost citrus phenotyping with comparatively high accuracy for accelerated breeding applications. Code is available at https://github.com/liangzhao2000/PhenoCitrus . • Low-cost phenotyping device: Dual cameras enable comprehensive fruit imaging. • Precise trait extraction: Multiple deep learning methods extract fruit traits. • Phenotypic profiling: Statistical analysis supports trait-based variety selection.

Why it matches plant phenotyping methods柑橘果実の形態形質を取得・抽出するハードウェア、画像解析、ソフトウェアを開発し、手動測定との相関で検証した中心的なフェノタイピング研究。

abstractAn integrated hardware–software pipeline is introduced to address this gap: (1) a custom imaging device for standardized top/side-view capture of the sample fruit, and (2) optimized computer vision algorithms using enhanced 3D Gaussian Splatting (3DGS) to extract 3D structural traits (surface area, volume) and 2D algorithms (e.g., Unet++ and YOLO variants) for 2D trait (width, length, oil cell number, peel/pulp color, peel thickness, and segment number) quantification.

OptiNet-B3: a lightweight explainable deep learning model for multiclass classification of fruit and leaf diseases. →

Early and accurate detection of diseases is very important for the health of crops and ensuring sustainable agricultural productivity. This paper proposes OptiNet-B3, a novel approach and an efficient deep model for the multiclass classification of fruit and leaf diseases for apples, bananas, and oranges. Through two diverse and comprehensive image datasets, the model performs well for both fruit 13,602 images and leaf 11,199 images classification. OptiNet-B3 optimizes learning in low computational budget by integrating Mish activation, Convolutional Block Attention Module (CBAM), Group Normalization, and knowledge distillation. Great care in preprocessing and augmenting data was taken to improve generalization. Comparison with state-of-the-art models-including DenseNet121, ResNet50, MobileNetV3, and InceptionV3-based models-reveals that OptiNet-B3 substantially outperforms in terms of accuracy, with 98.12% and 99.23% on the fruit and leaf datasets, respectively. Due to its light-weight architecture, real-time deployment for in-field diagnosis on mobile and edge devices is much more feasible. The results underscore the potential of explainable, AI-driven tools in transforming plant disease management practices.

Why it matches plant phenotyping methods果実・葉の画像から植物病害を分類するモデルを開発し、複数データセットと既存モデルとの比較で性能検証しているため、植物状態の画像ベース表現型推定が中心です。

abstractThis paper proposes OptiNet-B3, a novel approach and an efficient deep model for the multiclass classification of fruit and leaf diseases for apples, bananas, and oranges.

Unassailable citrus disease classification via multi-stage deep ensemble learning with vision transformers. →

To reduce losses from agriculture as well as enhance food security, we propose a three-stage deep ensemble for early citrus disease diagnosis from actual-field images of oranges (n = 2,240) as well as lemons (n = 208). To prevent leakage, augmentation is strictly enforced following splitting (70:30 stratified) following curation, normalisation, resizing by 224 × 224. Our contribution comes from combining state-of-the-art deep features by adding explicit texture priors. Namely, we add Local Binary Patterns (LBP) as well as Grey-Level Co-occurrence Matrix (GLCM) descriptors for micro-textures of lesions (e.g., stippling, scab rims around them, chlorosis encircled by veins) as well as second statistics (e.g., contrast, homogeneity, entropy) that CNNs/ViTs tend to discount by virtue of their small size coupled with variable-field data. These hand-crafted signals are z-score normalised as well as PCA-compressed for overfit protection as well as removal of collinearity then combined by deep embeddings. InceptionV3 (90% lemon) as well as DenseNet121 (93% orange) are the best of the five pretraining CNNs (ResNet50, DenseNet121, VGG16, InceptionV3, EfficientNetB0) that we test at Stage-1. The best CNNs are enlisted with a Vision Transformer (ViT) at Stage-2 for capture of long-range contextual capture improving upon Stage-1 by 98% (lemon) as well as 97% (orange). t-SNE confirms class separation while Stage-3 employs a multiclass SVM over the combined description that achieves 99% (lemon) while holding at 97% (orange) at another curation. The pipeline outperforms single-backbone variants, minimises variance while remaining lightweight enough for deployment, thus showing that LBP + GLCM texture priors compressed by PCA but combined by CNN/ViT features substantially enhance robustness plus generalisation for in-orchard citrus disease testing.

Why it matches plant phenotyping methods柑橘の実画像から病徴・病害状態を推定する深層学習画像解析パイプラインを開発・比較評価しており、植物フェノタイピング手法が中心である。

abstractwe propose a three-stage deep ensemble for early citrus disease diagnosis from actual-field images of oranges

A cross-cultivar hyperspectral framework for huanglongbing detection in citrus via wavelength optimization and deep learning. →

Huanglongbing (HLB) is a devastating disease that poses a serious threat to the global citrus industry. Due to its rapid spread and significant destructiveness, coupled with the lack of effective treatment methods, early and accurate detection is crucial for controlling disease spread and mitigating economic losses. Different citrus varieties exhibit significant differences in peel morphology and structure, leading to distinct hyperspectral reflection characteristics. This limits the applicability of traditional hyperspectral HLB detection methods across different varieties. Even when leaves with similar disease severity are detected, differences in reflection at sensitive wavelengths still exist, further limiting the adaptability of traditional hyperspectral detection methods across different varieties. To address this challenge, we propose a robust method for multi-variety HLB detection based on hyperspectral imaging. After data acquisition and preprocessing, the Successive Projections Algorithm (SPA) was used to extract characteristic wavelengths, and the Particle Swarm Optimization (PSO) algorithm was employed to identify wavelengths that remain consistent across different varieties. The statistical significance of these optimized wavelengths was verified using t-tests. The results showed that under specific conditions, there are significant differences in spectral responses among different varieties. This confirms that the selected wavelengths have cross-variety discrimination capability. Subsequently, the feature sets processed using SPA and PSO algorithms were used to train three classification models: Support Vector Machine (SVM), Multi-layer Perceptron (MLP), and a customized Convolutional Multi-scale Residual Network (CMR-CNN). The models were tested using the leave-one-out method. The PSO-optimized feature sets significantly improved model performance. In the setup where each variety was used as the test set in turn and the model was run ten times, model performance was expressed as the mean ± standard deviation (SD) of all variety test results. SVM accuracy increased from 89.38 % ± 0.81 % to 91.65 % ± 0.7 %; MLP accuracy increased from 89.58 % ± 1.19 % to 91.80 % ± 0.35 %; CMR-CNN accuracy increased from 90.83 % ± 0.43 % to 92.85 % ± 0.7 %. Due to its structural complexity and excellent feature extraction capability, the CMR-CNN model demonstrated the most outstanding diagnostic performance and showed great potential in plant disease diagnosis using hyperspectral imaging. This method establishes a universal HLB detection framework that does not require separate modeling for different citrus varieties.

Why it matches plant phenotyping methods柑橘葉の病徴状態を対象に、ハイパースペクトル画像の波長最適化と深層学習による品種横断的HLB検出フレームワークを開発・評価しており、植物病害表現型の取得・推定が中心である。

abstractwe propose a robust method for multi-variety HLB detection based on hyperspectral imaging.

Implementing transfer learning for citrus Huanglongbing disease detection across different datasets using neural network →

Citrus Huanglongbing (HLB) is highly contagious, and timely detection and removal of HLB-infected citrus trees is extremely important to prevent its spread. However, the robustness of optical imaging-based models remains limited by the variations in data due to different plant varieties, geospatial conditions, and data collection dates, etc. This study aimed to propose a method for robust HLB detection via transfer learning with multispectral-multicolor imaging. Four lightweight neural networks, namely Yolov7, Yolov7-tiny, Yolov4-tiny, and Mask-RCNN were introduced for citrus HLB disease detection across different datasets. Transfer learning on the Orah mandarin dataset was conducted using the Navel orange dataset for pre-training. The results showed that Mask-RCNN achieved the best performance with an mAP@0.5 of 91.65%. By replacing the backbone of Mask-RCNN with MobileNetV3-large, the model Mask-RCNNV3 was established, with an mAP@0.5 of 93.37% and then used for transfer learning for other datasts. Further optimizing the number of transferred layers and sample size, it revealed the most favorable sample size was 20 per class, and the mAP@0.5 gradually increased at the first 9 layers. Mask-RCNNV3 under the best transfer learning parameters, called Mask-RCNNV3_best, achieved the mAP@0.5 of 93.14% for Orah mandarin, 91.82% for Blood orange and 92.36% for Ponkan, respectively. Compared to the original Mask-RCNN model, the training parameters (Params) and GFLOPs were reduced by 82.95% and 96.57%, respectivley. It demonstrated that a limited amount of labeled data proved sufficient to achieve satisfactory performance across the tested cultivars and growing conditions. The FPS of the model was also improved by 4 times compared to Mask-RCNN, illustrating the potential of the model for edge deployment for practical applications. These findings would bridge the gap between research and practical implementation, reduce costly labeling for model training and provide practical tools for citrus growers to use.

Why it matches plant phenotyping methods柑橘HLB感染状态をマルチスペクトル画像から推定するニューラルネットワークを開発・比較・転移学習で検証しており、植物病害フェノタイピング手法が中心である。

abstractThis study aimed to propose a method for robust HLB detection via transfer learning with multispectral-multicolor imaging.

YOLO-Citrus: a lightweight and efficient model for citrus leaf disease detection in complex agricultural environments. →

Accurate and efficient detection of citrus leaf diseases is crucial for ensuring the quality and yield of global citrus production. However, many existing agricultural disease detection methods face significant challenges, including overlapping leaf occlusion, difficulty in identifying small lesions, and interference from complex backgrounds. These limitations often lead to reduced accuracy and efficiency of object detection. Moreover, current models generally necessitate significant computational resources and possess substantial model sizes, which restrict their practical applicability and operational convenience. To tackle these issues, this study presents a novel model named YOLO-Citrus. It is a lightweight and efficient YOLOv11-based model designed to enhance the precision of detection while simultaneously minimizing computational expenses and the size of the model. This makes it more suitable for practical agricultural applications. The proposed solution incorporates three major innovations: the C3K2-STA module, the ADown module, and the Wise-Inner-MPDIoU loss function. In particular, YOLO-Citrus utilizes Star-Triplet Attention by embedding Triplet Attention into the Star Block to enhance bottleneck performance in C3K2-STA. It also adopts the ADown module as a lightweight and effective downsampling strategy and introduces the Wise-Inner-MPDIoU loss to facilitate optimized bounding box regression and enhanced detection accuracy. These advancements enable high detection accuracy with substantially reduced computational requirements. The experimental results demonstrate that YOLO-Citrus attains 96.6% mAP@0.5, representing an improvement of 1.4 percentage points over the YOLOv11s baseline (95.2%). Furthermore, it reaches 81.6% mAP@0.5:0.95, i.e., an enhancement of 1.3 percentage points compared to the baseline value of 80.3%. The optimized model delivers considerable efficiency gains, with model size reduced by 25.0% from 19.2 MB to 14.4 MB and computational cost decreased by 20.2% from 21.3 to 17.0 GFlops. Comparative analysis has confirmed that YOLO-Citrus performs better than other models in terms of comprehensive detection capability. These performance enhancements validate the model's effectiveness in real-world orchard conditions, offering practical solutions for early disease detection, precision treatment, and yield protection in citrus cultivation.

Why it matches plant phenotyping methods柑橘葉の病害を画像から検出するYOLOモデルの開発・性能評価が中心であり、植物の病害状態を直接推定するフェノタイピング手法に該当する。

abstractThe experimental results demonstrate that YOLO-Citrus attains 96.6% mAP@0.5

External defect detection of Orah mandarin based on a non-brightness correction algorithm. →

External defect detection is a crucial step in Orah mandarin citrus grading. However, in existing defect detection algorithms by image processing, Orah mandarin surfaces exhibit characteristics such as higher brightness at the center, lower brightness at the edges, and uneven brightness distribution in images. Although traditional brightness correction algorithms can solve these issues, they suffer from limitations including prolonged processing time, high computational complexity, and elevated false detection rates. To address these shortcomings, this work proposes a non-brightness correction algorithm to enhance the speed and accuracy of Orah mandarin external defect detection. The proposed algorithm divides Orah mandarin images into multiple equal-sized regions and performs threshold segmentation sequentially using a sliding window matching the region size. A sliding window size of 100 × 100 pixels was chosen because it offers a balanced trade-off between detection precision and computational efficiency, allowing the algorithm to detect both large and subtle defects effectively while maintaining fast processing speed. First, the histogram statistical method categorizes the current sliding window region into three types, and a dedicated defect detection algorithm applies adaptive thresholding to each type. Next, the threshold-segmented regions are merged, while the fruit stem area is excluded by combining circularity and hue features. Finally, morphological operations eliminate noise to obtain complete defect segmentation results. Experimental results demonstrate that with a sliding window size of 100 × 100 pixels, the algorithm achieves rapid external defect detection at 85.3 ms per fruit and a 97.5% defect recognition rate, offering a novel approach for fruit surface defect detection. This performance is consistent across different defect types, though the algorithm performed best for point-like defects, such as thrips scarring and canker spots, where clear, localized defects were more easily detected. For blocky rot defects, such as sunburn, the algorithm exhibited a slightly lower recognition rate, particularly in areas where the defect was less distinct and more integrated with the fruit's surface. These findings suggest that the algorithm is effective for a range of defect types but may require further refinement to handle more complex or overlapping defects.

Why it matches plant phenotyping methods柑橘果実表面の外観欠陥という植物器官の状態を、画像処理・適応閾値処理・形態学的処理で抽出する手法を開発し、速度と認識率を評価しており、フェノタイピング手法が中心です。

abstractTo address these shortcomings, this work proposes a non-brightness correction algorithm to enhance the speed and accuracy of Orah mandarin external defect detection.

Quantifying high-temperature-induced reproductive growth imbalance in citrus at anthesis: Insights from the CF-ASPM model →

Global climate change-induced environmental stress poses critical challenges to the stable development of economic crops such as citrus. High temperatures (HTs) at anthesis may cause poor pollination and excessive flower/fruit drop, seriously affecting fruit yield and quality. To comprehensively analyze the developmental dynamics and morphological responses of citrus to HT stress at anthesis, methods for precise whole-flower phenotypic extraction and stamen state classification were developed. A citrus flower automatic segmentation and phenotypic quantitative model (CF-ASPM) that combines the pre-trained Segment Anything Model (SAM) with a lightweight classification module was constructed to accurately identify and quantify key citrus flower structures. Phenotypic parameter extraction correlation coefficients were 0.90–0.98. A few-shot stamen classification method was also designed using a pre-segmentation strategy and differential features, and its classification accuracy was 96.39%. Experiments with Ehime mandarin were conducted to analyze dynamic citrus floral organ changes at different temperatures and the underlying physiological mechanisms. The results showed that citrus exhibits a distinct reproductive priority strategy under HTs. Floral organ growth is inhibited, blooming is accelerated, and an asynchronous compensation mechanism occurs between male and female organs. HTs accelerated flower aging and caused developmental imbalances in the ovary and nectar disc. This may lead to increased flower and fruit drop and altered fruit shape. This study revealed the reproductive priority strategy and growth imbalance of citrus floral organs under HTs using the CF-ASPM model. It provides important data for further exploring the molecular mechanisms and management strategies of HT stress.

Why it matches plant phenotyping methods柑橘花器官の自動セグメンテーション、形質抽出、雄蕊状態分類法を開発し、精度検証したことが研究の中心であるため。

abstractmethods for precise whole-flower phenotypic extraction and stamen state classification were developed.

Reinforcement Intelligence for Spectral Enhancement (RISE): A novel feature extraction method for hyperspectral prediction of sugar content in Citrus reticulata 'Chun Jian' →

This study proposes a reinforcement learning-based hyperspectral feature band selection method, Reinforcement Intelligence for Spectral Enhancement (RISE), for the non-destructive detection of sugar content in citrus. The band selection problem is modelled as a Markov decision process, and an optimal feature optimisation strategy is learned through a deep Q network. A total of 120 citrus pulp samples with sugar content ranging from 6.40 to 10.81 °Brix were collected. Hyperspectral data (388.34–1036.34 nm) containing 256 continuous bands were collected in the experiment, and compared with the traditional CARS (9 bands selected) and BOSS (10 bands selected) algorithms. The results show that the RISE algorithm selected 19 characteristic bands that obtained the best prediction performance (R² = 0.84, RPD = 2.51) on the PLSR model, and maintained consistent performance across multiple prediction models including SVR (R² = 0.85, RPD = 2.57), Random Forest (R² = 0.84, RPD = 2.47) and XGBoost (R² = 0.84, RPD = 2.53). The visualization of the spatial distribution of sugar content in citrus fruits based on the RISE algorithm revealed a gradient distribution feature that decreases from the outside to the inside. The study confirms the application potential of the RISE algorithm in non-destructive testing of agricultural product quality and provides a new technical path for hyperspectral imaging technology in agriculture.

Why it matches plant phenotyping methods柑橘果实糖含量是植物器官性状，研究核心是开发并比较基于高光谱数据的特征波段选择与无损性状预测方法，而非例行测量。

abstractThis study proposes a reinforcement learning-based hyperspectral feature band selection method, Reinforcement Intelligence for Spectral Enhancement (RISE), for the non-destructive detection of sugar content in citrus.

Agrosense: Accelerating precision orchard management through an AI-enabled monitoring system →

PURPOSE: Efficient orchard management requires high-throughput phenotyping technologies to assist growers in crop monitoring and decision-making. This study presents Agrosense, an advanced artificial intelligence (AI) powered sensing system designed for real-time phenotypic data collection in orchards, addressing the limitations of traditional manual methods. METHODS: Agrosense integrates four RGB-D cameras with a Jetson Xavier microprocessor to collect high-resolution data and perform tree crop counting, canopy density classification, and tree height estimation. A citrus orchard served as a case study, where 337 trees were imaged to train and validate AI models. YOLOv8 was employed for object detection and classification tasks, while five methods were tested for estimating tree height. RESULTS: The YOLOv8 model achieved a mean average precision (mAP) of 0.977 for tree trunkdetection and 0.974 for canopy density classification. In field testing on 157 citrus trees, the system achieved 95% accuracy for tree trunk detection and 94% accuracy for canopy density classification, with only 11 misclassifications. The best-performing method for height estimation achieved a mean absolute percentage error (MAPE) of 8.53%. Agrosense completed phenotyping tasks in 398 s, a 515% speed improvement over manual methods (2,446 s). CONCLUSION: Agrosense effectively supports precision orchard management by automating key phenotyping tasks with high accuracy and efficiency. The system significantly reduces data collection time and improves consistency. Future work will focus on algorithm refinement and adaptation to other tree crops to broaden the system’s utility in precision agriculture.

Why it matches plant phenotyping methodsRGB-DカメラとAIを統合した果樹フェノタイピングシステムを開発・検証し、樹冠密度や樹高などの形質を定量化しているため、方法が研究の中心である。

abstractThis study presents Agrosense, an advanced artificial intelligence (AI) powered sensing system designed for real-time phenotypic data collection in orchards

Vision Transformer-Capsule Network for Orange Quality Inspection: Ripeness and Black Spot Disease Detection →

The accurate classification of orange fruit ripeness and early detection of citrus black spot disease are crucial for optimizing harvest timing, ensuring post-harvest quality, and minimizing economic losses. Traditional manual inspection methods are subjective and inconsistent, necessitating automated deep learning techniques. This study proposes a ViT-CapsNet hybrid model, integrating Vision Transformers (ViT) for global feature extraction with Capsule Networks (CapsNet) for spatial hierarchy preservation, leading to superior classification accuracy, segmentation performance, and robustness. The dataset comprises 400 images Unripe (100), Half-Ripe (100), Ripe (100), and Infected (100) sourced from Hugging Face. Data augmentation, including rotation, brightness adjustment, flipping, and CutMix, expanded the dataset to 1200 images, improving generalization. The ViT-CapsNet model achieves 95.37% training accuracy and 96.12% validation accuracy, outperforming ViT-only (88.45%), CapsNet-only (86.39%), CNN (89.72%), and ViT-CNN hybrid (90.81%). The F1-score per class is Unripe: 0.88, Half-Ripe: 0.85, Ripe: 0.94, Infected: 0.92. The Intersection over Union (IoU) score (0.70) and dice coefficient (0.82) surpass CNN’s 0.65 and 0.78, respectively. The precision-recall under the curve (AUC) is 0.93, exceeding ResNet-50 (0.86), MobileNetV2 (0.84), and CNN (0.88). The model maintains 94.21% accuracy under normal conditions, dropping to 85.37% (Gaussian blur), 80.23% (JPEG compression), and 78.64% (occlusions). The out-of-distribution (OOD) detection score is 0.84, and inference time is 0.05 s. Training used AdamW optimizer (learning rate: 0.0001, batch size: 32, 60 epochs), achieving 0.301 validation loss. The proposed ViT-CapsNet model is a scalable and efficient solution for real-time agricultural automation.

Why it matches plant phenotyping methodsオレンジ果実の熟度および黒点病状態を画像から推定する深層学習手法を開発し、複数モデルとの精度比較、頑健性、OOD性能、推論時間を評価しており、植物状態の取得・判定法が中心である。

abstractThis study proposes a ViT-CapsNet hybrid model, integrating Vision Transformers (ViT) for global feature extraction with Capsule Networks (CapsNet) for spatial hierarchy preservation, leading to superior classification accuracy, segmentation performance, and robustness.

Deriving Early Citrus Fruit Yield Estimation by Combining Multiple Growing Period Data and Improved YOLOv8 Modeling. →

Early crop yield prediction is a major challenge in precision agriculture, and efficient and rapid yield prediction is highly important for sustainable fruit production. The accurate detection of major fruit characteristics, including flowering, green fruiting, and ripening stages, is crucial for early yield estimation. Currently, most crop yield estimation studies based on the YOLO model are only conducted during a single stage of maturity. Combining multi-growth period data for crop analysis is of great significance for crop growth detection and early yield estimation. In this study, a new network model, YOLOv8-RL, was proposed using citrus multigrowth period characteristics as a data source. A citrus yield estimation model was constructed and validated by combining network identification counts with manual field counts. Compared with YOLOv8, the number of parameters of the improved network is reduced by 50.7%, the number of floating-point operations is decreased by 49.4%, and the size of the model is only 3.2 MB. In the test set, the average recognition rate of citrus flowers, green fruits, and orange fruits was 95.6%, the mAP@.5 was 94.6%, the FPS value was 123.1, and the inference time was only 2.3 milliseconds. This provides a reference for the design of lightweight networks and offers the possibility of deployment on embedded devices with limited computational resources. The two estimation models constructed on the basis of the new network had coefficients of determination R 2 values of 0.91992 and 0.95639, respectively, with a prediction error rate of 6.96% for citrus green fruits and an average error rate of 3.71% for orange fruits. Compared with network counting, the yield estimation model had a low error rate and high accuracy, which provided a theoretical basis and technical support for the early prediction of fruit yield in complex environments.

Why it matches plant phenotyping methods柑橘の花・果実を画像認識して収量を推定するYOLOv8改良モデルとワークフローを開発・検証しており、植物形質取得法が中心的です。

abstractIn this study, a new network model, YOLOv8-RL, was proposed using citrus multigrowth period characteristics as a data source.

Trends and Future Challenges in Image Analysis for Digital Forecasting of Jeju Citrus Production →

This review paper focuses on the digital transformation of yield prediction for the sustainable development of the Jeju citrus industry, emphasizing current applications and future challenges of image analysis technologies. Accurate yield prediction is essential for stabilizing farm income, improving distribution efficiency, balancing supply and demand, and optimizing cultivation strategies. However, traditional statistics-based approaches are limited by climate change, cultivation area fluctuations, and labor shortages. In this context, Al-driven digital technologies —especially non-invasive image analysis —have emerged as promising alternatives. The paper provides an in-depth overview of image analysis applications in two key areas: fruit detection and counting, and fruit size and growth prediction. Notably, deep learning-based object detection models (e.g., YOLO, Faster R-CNN) and 3D reconstruction technologies have improved prediction accuracy. Integrating auxiliary data, such as maturity and quality indicators, is also discussed. Despite these advancements, challenges remain for real-world implementation. These include data collection under varied environments, model robustness (especially against occlusion), and the need for real-time processing and user-friendly system design. Future research should prioritize integrating heterogeneous data — including weather and soil — long-term time-series learning, and developing cost-effective, high-efficiency solutions. These efforts are expected to enhance the accuracy and reliability of citrus yield predictions, driving the digital transformation and sustainable future of the Jeju citrus industry.

Why it matches plant phenotyping methods画像解析による柑橘果実の検出・計数、サイズ・成長推定を中心にレビューしており、植物形質取得手法が主要内容である。

abstractThe paper provides an in-depth overview of image analysis applications in two key areas: fruit detection and counting, and fruit size and growth prediction.

Advances in Precision Farming: a contribute for estimating crop health and water stress by comparing UAV Multispectral and Thermal Imagery →

Abstract. Current climate change is largely due to the continuing increase in the anthropogenic greenhouse effect, with major environmental repercussions, especially in agriculture. The increase of global warming, salinity of water resources and frequency of extreme weather events has devastating consequences on the primary sector, in particular on the photosynthetic activity of crops and, therefore, their agricultural yield. The current climate crisis, in fact, leads to an increase in water requirements, the proliferation of weeds, and the depletion of nutrients in the soil, necessitating the massive use of fertilisers, herbicides and pesticides, which, in turn, trigger substantial alterations in ecosystem balances. In response to these critical issues, precision agriculture (PA) constitutes a data-driven approach based on the interpretation of multispectral and thermal datasets obtained by different remote sensing techniques and the use of latest-generation sensors to recognise the state of health of crops and, therefore, optimise agricultural production with a more rational and sustainable management of resources.This paper presents the results of a survey campaign carried out in October 2023 on two citrus fields located in south-eastern Sicily (Italy) to highlight the health status of crops just before the harvesting period. By using multispectral and thermal sensors installed on a drone, different vegetation indices have been calculated to identify, in each field, the areas with the highest photosynthetic activity and the zones characterised by a lack of water or other nutrients, on which targeted agronomic interventions should be planned as a priority.

Why it matches plant phenotyping methodsUAVマルチスペクトル・熱画像を用いて、柑橘作物の健康状態、光合成活性、水ストレスを推定するセンシング手法の適用が中心であり、単なるルーチン測定ではない。

titleestimating crop health and water stress by comparing UAV Multispectral and Thermal Imagery

YOLOv8-Scm: an improved model for citrus fruit sunburn identification and classification in complex natural scenes. →

Citrus ranks among the most widely cultivated and economically vital fruit crops globally, with southern China being a major production area. In recent years, global warming has intensified extreme weather events, such as prolonged high temperature and strong solar radiation, posing increasing risks to citrus production,leading to significant economic losses. Existing identification methods struggle with accuracy and generalization in complex environments, limiting their real-time application. This study presents an improved, lightweight citrus sunburn recognition model, YOLOv8-Scm, based on the YOLOv8n architecture. Three key enhancements are introduced: (1) DSConv module replaces the standard convolution for a more efficient and lightweight design, (2) Global Attention Mechanism (GAM) improves feature extraction for multi-scale and occluded targets, and (3) EIoU loss function enhances detection precision and generalization. The YOLOv8-Scm model achieves improvements of 2.0% in mAP50 and 1.5% in Precision over the original YOLOv8n, with only a slight increase in computational parameters (0.182M). The model's Recall rate decreases minimally by 0.01%. Compared to other models like SSD, Faster R-CNN, YOLOv5n, YOLOv7-tiny, YOLOv8n, and YOLOv10n, YOLOv8-Scm outperforms in mAP50, Precision, and Recall, and is significantly more efficient in terms of computational parameters. Specifically, the model achieves a mAP50 of 92.7%, a Precision of 86.6%, and a Recall of 87.2%. These results validate the model's superior capability in accurately detecting citrus sunburn across diverse and challenging natural scenarios. YOLOv8-Scm enables accurate, real-time citrus sunburn monitoring, providing strong technical support for smart orchard management and practical deployment.

Why it matches plant phenotyping methods柑橘果実のサンバーンという植物の病徴・状態を画像から検出・分類するモデルを開発し、複数モデルとの性能比較で技術的に検証しているため、植物フェノタイピング手法が中心である。

abstractThis study presents an improved, lightweight citrus sunburn recognition model, YOLOv8-Scm, based on the YOLOv8n architecture.

A Multi-kernel CNN model with attention mechanism for classification of citrus plants diseases. →

One of the primary challenges leading to a significant reduction in agricultural production is the prevalence of diseases affecting citrus plants. Prevention and monitoring the spread of citrus plant diseases is crucial for maintaining citrus production. This decrease in productivity adversely affects the overall economy. The essential step for enhancing the quality of fruit production and promoting economic growth involves the classification and identification of leaf diseases in the early stage. In this work, a multi-kernel CNN model with attention mechanism is used for classification of citrus plants diseases is proposed. Initially, the input image is pre-processed for resizing the images as the images are obtained from different datasets. After resizing the image, the feature extraction process is carried out by the pretrained convolutional neural networks. In the next step, the two attention mechanisms multi kernel channel attention and spatial attention is used. These two attention mechanisms are used for obtaining spatial and channel attention feature maps. Finally, the classification process is carried out to classify the normal and diseased cases. The test accuracy results shows that our model surpasses the other models in terms of its classification performance.

Why it matches plant phenotyping methods柑橘葉画像から病害状態を分類するCNN手法が研究の中心であり、植物の病徴・状態を画像から推定するため、植物フェノタイピング手法として含める。

abstractIn this work, a multi-kernel CNN model with attention mechanism is used for classification of citrus plants diseases is proposed.

Multiclass semantic segmentation for prime disease detection with severity level identification in Citrus plant leaves →

Agriculture provides the basics for producing food, driving economic growth, and maintaining environmental sustainability. On the other hand, plant diseases have the potential to reduce crop productivity and raise expenses, posing a risk to food security and the incomes of farmers. Citrus plants, recognized for their nutritional benefits and economic significance, are especially vulnerable to diseases such as citrus greening, Black spot, and Citrus canker. Due to technological advancements, image processing and Deep learning algorithms can now detect and classify plant diseases early on, which assists in preserving crop health and productivity. The proposed work enables farmers to identify and visualize multiple diseases affecting citrus plants. This study proposes an efficient model to detect multiple citrus diseases (canker, black spot, and greening) that may co-occur on the same leaf. It is achieved using the RSL (Residual Squeeze & Excitation LeakyRelu) Linked-TransNet multiclass segmentation model. The proposed model stands out in its ability to address major limitations in existing models, including spatial inconsistency, loss of fine disease boundaries, and inadequate feature representation. The significance of this proposed RSL Linked-Transnet model lies in its integration of hierarchical feature extraction, global context modeling via transformers, and precise feature reconstruction, ensuring superior segmentation accuracy and robustness. The results of the proposed RSL Linked-TransNet architecture reveal average values of 0.9755 for accuracy, 0.0660 for loss, 0.9779 for precision, 0.9738 for recall, and 0.9308 for IoU. Additionally, the model achieves a mean F1 score of 0.7173 and a mean IoU of 0.7567 for each disease class in images from the test dataset. The segmentation results are further utilized to identify the prime disease affecting the leaves and evaluate disease severity using the prime disease classification and severity detection algorithm.

Why it matches plant phenotyping methods柑橘葉画像から病斑をセグメンテーションし、主要病害と重症度を推定する手法の開発が中心であり、植物の病害状態を直接測定するため、植物フェノタイピング手法として適格。

abstractThis study proposes an efficient model to detect multiple citrus diseases (canker, black spot, and greening) that may co-occur on the same leaf.

Multi-component image analysis for citrus disease detection using convolutional neural networks →

Citrus crops are susceptible to diseases such as Black Spot, Canker, and Greening, which significantly harm both the leaves and the fruits, ultimately reducing overall yield. Traditional visual inspection methods for identifying these diseases are labour-intensive and prone to inaccuracies. The present research proposes a deep learning approach utilizing Convolutional Neural Networks (CNNs) to overcome the limitations of manual inspection. Moreover, it introduces the utilization of combined visual features of citrus leaves and fruits for enhanced disease classification. The proposed multi-component approach demonstrates superior classification performance, achieving more accurate results than single-component-based classifications. A dataset comprising 12,000 images, distributed across leaves, fruits, and their merged forms, was used for training, validation, and testing. Three CNN models were developed and evaluated: Leaf-Trained, Fruit-Trained, and Multiple Component-Trained CNNs. Performance was assessed using metrics such as accuracy, precision, recall, and F1-score, including their macro values, focusing on model generalization across different input types. The Multiple Component-Trained CNN outperformed the other models, achieving a validation accuracy of 97.75%, followed by the Leaf-Trained CNN at 95.50%. During testing, it also demonstrated superior performance across all input types, with accuracies of 94.75% on the leaf dataset, 92.87% on the fruit dataset, and 96.62% on the merged dataset. The results indicate that Black Spot is the most accurately classified disease, while Canker and Greening are less accurately classified. These findings highlight the potential of integrating various components of plants for enhanced disease classifications.

Why it matches plant phenotyping methods柑橘の葉・果実画像から病害状態を推定するCNN手法を開発・比較し、複数入力による分類性能を検証しているため、植物表現型取得手法が中心である。

abstractThe present research proposes a deep learning approach utilizing Convolutional Neural Networks (CNNs) to overcome the limitations of manual inspection.

On the Usage of Deep Learning Techniques for Unmanned Aerial Vehicle-Based Citrus Crop Health Assessment →

This work proposes an end-to-end solution for leaf segmentation, disease detection, and damage quantification, specifically focusing on citrus crops. The primary motivation behind this research is to enable the early detection of phytosanitary problems, which directly impact the productivity and profitability of Spanish and Portuguese agricultural developments, while ensuring environmentally safe management practices. It integrates an onboard computing module for Unmanned Aerial Vehicles (UAVs) using a Raspberry Pi 4 with Global Positioning System (GPS) and camera modules, allowing the real-time geolocation of images in citrus croplands. To address the lack of public data, a comprehensive database was created and manually labelled at the pixel level to provide accurate training data for a deep learning approach. To reduce annotation effort, we developed a custom automation algorithm for pixel-wise labelling in complex natural backgrounds. A SegNet architecture with a Visual Geometry Group 16 (VGG16) backbone was trained for the semantic, pixel-wise segmentation of citrus foliage. The model was successfully integrated as a modular component within a broader system architecture and was tested with UAV-acquired images, demonstrating accurate disease detection and quantification, even under varied conditions. The developed system provides a robust tool for the efficient monitoring of citrus crops in precision agriculture.

Why it matches plant phenotyping methods柑橘葉のセグメンテーション、病害検出・被害量定を行う画像解析とUAV搭載システムを開発・評価しており、植物の病害状態を直接推定する方法が研究の中心である。

abstractThis work proposes an end-to-end solution for leaf segmentation, disease detection, and damage quantification, specifically focusing on citrus crops.

YOLOv10-LGDA: An Improved Algorithm for Defect Detection in Citrus Fruits Across Diverse Backgrounds. →

Citrus diseases can lead to surface defects on citrus fruits, adversely affecting their quality. This study aims to accurately identify citrus defects against varying backgrounds by focusing on four types of diseases: citrus black spot, citrus canker, citrus greening, and citrus melanose. We propose an improved YOLOv10-based disease detection method that replaces the traditional convolutional layers in the Backbone network with LDConv to enhance feature extraction capabilities. Additionally, we introduce the GFPN module to strengthen multi-scale information interaction through cross-scale feature fusion, thereby improving detection accuracy for small-target diseases. The incorporation of the DAT mechanism is designed to achieve higher efficiency and accuracy in handling complex visual tasks. Furthermore, we integrate the AFPN module to enhance the model's detection capability for targets of varying scales. Lastly, we employ the Slide Loss function to adaptively adjust sample weights, focusing on hard-to-detect samples such as blurred features and subtle lesions in citrus disease images, effectively alleviating issues related to sample imbalance. The experimental results indicate that the enhanced model YOLOv10-LGDA achieves impressive performance metrics in citrus disease detection, with accuracy, recall, mAP@50, and mAP@50:95 rates of 98.7%, 95.9%, 97.7%, and 94%, respectively. These results represent improvements of 4.2%, 3.8%, 4.5%, and 2.4% compared to the original YOLOv10 model. Furthermore, when compared to various other object detection algorithms, YOLOv10-LGDA demonstrates superior recognition accuracy, facilitating precise identification of citrus diseases. This advancement provides substantial technical support for enhancing the quality of citrus fruit and ensuring the sustainable development of the industry.

Why it matches plant phenotyping methods柑橘果実の病斑・表面欠陥を画像から検出する改良YOLO手法の開発と性能評価が中心であり、植物器官の病害状態を推定する画像ベース表現型計測に該当する。

abstractWe propose an improved YOLOv10-based disease detection method

YOLOV8-CMS: a high-accuracy deep learning model for automated citrus leaf disease classification and grading. →

Background Citrus leaf diseases significantly affect production efficiency and fruit quality in the citrus industry. To effectively identify and classify citrus leaf diseases, this study proposed a classification approach leveraging deep learning techniques (YOLOV8 equipped with CSPPC, MultiDimen, SpatialConv, YOLOV8-CMS). Additionally, a segmentation method was utilized to extract leaf and lesion areas for disease severity grading based on their pixel ratio. Results By collecting and preprocessing a citrus leaf image dataset, the YOLOV8-CMS model was trained for disease classification. The model integrated MultiDimen attention, SpatialConv, and the CSPPC module to enhance performance. Furthermore, a segmentation approach was applied to precisely segment both leaf and lesion areas, enabling a quantitative assessment of disease severity. To verify the effectiveness of the proposed approach, multiple YOLO-based architectures, including different YOLOV8 series models, YOLOV5, and YOLOV3, were compared and analyzed. Results demonstrated that the proposed method achieved outstanding performance in citrus leaf disease classification, with an mAP50 of 98.2% in distinguishing healthy and diseased leaves and an accuracy of 97.9% in multi-class disease classification tasks. Conclusions The proposed YOLOV8-CMS model outperformed traditional methods in citrus leaf disease classification, while the segmentation-based approach enabled an accurate and quantitative assessment of disease severity. These findings highlighted the potential of deep learning in precision agriculture, contributing to more effective disease management in citrus production.

Why it matches plant phenotyping methods柑橘葉の病害分類と、葉・病斑の画像分割による病害重症度の定量化手法が研究の中心であり、植物状態の表現型を直接推定・検証している。

abstracta segmentation method was utilized to extract leaf and lesion areas for disease severity grading based on their pixel ratio.

Design and Development of a Precision Spraying Control System for Orchards Based on Machine Vision Detection. →

Precision spraying technology has attracted increasing attention in orchard production management. Traditional chemical pesticide application relies on subjective judgment, leading to fluctuations in pesticide usage, low application efficiency, and environmental pollution. This study proposes a machine vision-based precision spraying control system for orchards. First, a canopy leaf wall area calculation method was developed based on a multi-iteration GrabCut image segmentation algorithm, and a spray volume calculation model was established. Next, a fuzzy adaptive control algorithm based on an extended state observer (ESO) was proposed, along with the design of flow and pressure controllers. Finally, the precision spraying system's performance tests were conducted in laboratory and field environments. The indoor experiments consisted of three test sets, each involving six citrus trees, totaling eighteen trees arranged in two staggered rows, with an interrow spacing of 3.4 m and an intra-row spacing of 2.5 m; the nozzle was positioned approximately 1.3 m from the canopy surface. Similarly, the field experiments included three test sets, each selecting eight citrus trees, totaling twenty-four trees, with an average height of approximately 1.5 m and a row spacing of 3 m, representing a typical orchard environment for performance validation. Experimental results demonstrated that the system reduced spray volume by 59.73% compared to continuous spraying, by 30.24% compared to PID control, and by 19.19% compared to traditional fuzzy control; meanwhile, the pesticide utilization efficiency increased by 61.42%, 26.8%, and 19.54%, respectively. The findings of this study provide a novel technical approach to improving agricultural production efficiency, enhancing fruit quality, reducing pesticide use, and promoting environmental protection, demonstrating significant application value.

Why it matches plant phenotyping methods画像分割により果樹の樹冠葉壁面積という植物形態形質を抽出し、その値に基づく散布量制御システムを開発・検証しており、形質取得手法が中心的です。

abstracta canopy leaf wall area calculation method was developed based on a multi-iteration GrabCut image segmentation algorithm

Identification of yellow vein clearing disease in lemons based on hyperspectral imaging and deep learning. →

Hyperspectral imaging (HSI) technology has great potential for the efficient and accurate detection of plant diseases. To date, no studies have reported the identification of yellow vein clearing disease (YVCD) in lemon plants by using hyperspectral imaging. A major challenge in leveraging HSI for rapid disease diagnosis lies in efficiently processing high-dimensional data without compromising classification accuracy. In this study, hyperspectral feature extraction is optimized by introducing a novel hybrid 3D-2D-LcNet architecture combined with three-dimensional (3D) and two-dimensional (2D) convolutional layers-a methodological advancement over conventional single-mode CNNs. The competitive adaptive reweighted sampling (CARS) and successive projection algorithm (SPA) were utilized to reduce the dimensionality of hyperspectral images and select the feature wavelengths for YVCD diagnosis. The spectra and hyperspectral images retrieved through feature wavelength selection were separately employed for the modeling process by using machine learning algorithms and convolutional neural network algorithms (CNN). Machine learning algorithms (such as support vector machine and partial least squares discriminant analysis) and convolutional neural network algorithms (CNN) (including 3D-ShuffleNetV2, 2D-LcNet and 2D-ShuffleNetV2) were utilized for comparison analysis. The results showed that CNN-based models have achieved an accuracy ranging from 93.90% to 97.35%, significantly outperforming machine learning approaches (ranging from 68.83% to 93.52%). Notably, the hybrid 3D-2D-LcNet has achieved the highest accuracy of 97.35% (CARS) and 96.86% (SPA), while reducing computational costs compared to 3D-CNNs. These findings suggest that hybrid 3D-2D-LcNet effectively balances computational complexity with feature extraction efficacy and robustness when handling spectral data of different wavelengths. Overall, this study offers insights into the rapidly processing hyperspectral images, thus presenting a promising method.

Why it matches plant phenotyping methodsレモンの病徴を対象に、ハイパースペクトル画像の特徴抽出・波長選択・深層学習モデルを開発し、精度比較で検証しているため、植物病害状態のフェノタイピング手法が中心である。

abstractThe competitive adaptive reweighted sampling (CARS) and successive projection algorithm (SPA) were utilized to reduce the dimensionality of hyperspectral images and select the feature wavelengths for YVCD diagnosis.

Enhancing Citrus Plant Health through the Application of Image Processing Techniques for Disease Detection →

The foremost task in agriculture is the decisive identification of citrus plants and the timely identification of diseases in the plants with the aim of improving the quality of crops and the yield. In this work, a machine learning algorithm focuses on image processing of citrus to solve issues that are significant and cause concern in agriculture. This work focus on the machine learning models like VGG 19 and VGG 16. In addition, dataset curation, data augmentation and various other methods were employed. The dataset used in this research is a composed one which is recorded in a comprehensive manner including the data of both the affected and healthy pieces of citrus fruits. The ensemble model utilised here to ensure the improvement of trained datasets. Reviewing the research on machine learning models indicates a possibility for accurate classification of the fruits and disease detection models of the fruit. The three contenders performed admirably, with VGG 19 dominating with 95.5% accuracy. In second place was CNN with 93.4% and VGG 16 trailing at 91.2%. Such models are recognisable, because they perform well in agricultural environments, thanks to their precision, recall, and F1 scores, which are all balanced properly. The models’ capacity to lessen the number of false alarms and misses is further assessed with the use of confusion matrices, which are of utmost importance in disease control. New developments in early disease diagnosis and detection of citrus fruits in agriculture may greatly enhance the health and productivity of crops. This research can be critical in increasing agricultural productivity while ensuring the environmental sustainability and health of growers and citrus crops in the long run.

Why it matches plant phenotyping methods柑橘の画像から健全・罹病状態を分類する機械学習・画像処理手法が研究の中心であり、植物病害状態のフェノタイピングに該当する。

titleEnhancing Citrus Plant Health through the Application of Image Processing Techniques for Disease Detection

Phenotyping drought stress tolerance in citrus rootstocks using high-throughput imaging and physio-biochemical techniques. →

Background Drought stress, the most prevalent abiotic stress, has a significant effect on citrus production worldwide. The differential mechanisms to overcome the drought stress has been reported in citrus rootstock genotypes. This study evaluated nine citrus rootstock genotypes, including indigenous rough lemon variants, for drought tolerance. The genotypes were subjected to well-watered, drought stress, and re-watering conditions to assess morphological, physiological, and biochemical responses. High-throughput imaging techniques were employed to non-destructively assess chlorophyll fluorescence, digital leaf area, and plant tissue water content during drought stress. Results For rapid and accurate screening of rootstocks, phenomics and physio-biochemical tools were used to know morpho-physiological responses to drought. Citrus rootstock genotype X639 demonstrated superior performance under drought stress conditions. It maintained the highest growth in terms of relative shoot increment (8.09%), number of leaves (79.00), and specific leaf area (62.45 cm 2 g -1 ). X639 also excelled in root morphological parameters, including root length, projected area, diameter, surface area, volume, and number of tips, forks, and crossings. Trifoliate hybrids X639 and Troyer citrange exhibited larger stomata (54.73 and 43.82 µm 2 ) compared to mono-foliate species, with minimal impact of drought on stomatal pore area. X639 maintained the highest relative water content, membrane and chlorophyll stability indices, leaf gas exchange parameters, and antioxidant enzyme activity. RLC-1 and RLC-4 genotypes showed pronounced accumulation of leaf proline and antioxidant enzymes during drought, contributing to better recovery after re-watering. Conclusion In this study, Cleopatra mandarin, Grambhiri, and RLC-2 were identified as drought-susceptible rootstocks based on their responses. Rootstock genotypes X639 and RLC-4 proven a superior drought-tolerant genotypes. Their robust root system enables efficient water uptake and the maintenance of water relations during drought stress. The drought tolerance of X639 was evidenced by its ability to maintain plant tissue moisture, membrane and chlorophyll stability, and higher photosystem II efficiency. High-throughput imaging techniques have proven effective in rapidly assessing and differentiating drought-tolerant and drought-susceptible citrus rootstocks based on their photosystem- II efficiency, leaf area, and tissue water content during induced drought stress. These findings will contribute to the selection and development of drought-tolerant citrus rootstocks to improve citrus production under water-limited conditions.

Why it matches plant phenotyping methods高スループット画像法を用いて葉面積、クロロフィル蛍光、組織含水量を非破壊測定し、乾燥耐性スクリーニングに適用・評価しており、表現型取得法が中心的です。

abstractHigh-throughput imaging techniques were employed to non-destructively assess chlorophyll fluorescence, digital leaf area, and plant tissue water content during drought stress.

In situ nondestructive identification of citrus fruit ripeness via hyperspectral imaging technology. →

Rapid and accurate assessment of the citrus ripening stage in the field is important for determining harvest timing and improving industrial economic efficiency; however, the lack of effective nondestructive detection methods in the current orchard leads to flaws in ripening stage assessment, which affects harvesting decisions. To solve this problem, this study utilized hyperspectral technology to collect data from 22 fruit trees in an orchard (in the range of 400-1000 nm) and explored the effectiveness of five regions of interest selection methods (x-axis, y-axis, four-quadrant, threshold segmentation, and raw) for the delineation of the citrus ripening stage. The data quality was enhanced via wavelet transform (WT)-multiple scattering correction (MSC) preprocessing, and the effective wavelengths were extracted via the successive projections algorithm (SPA). On the basis of these wavelengths, backpropagation neural network (BP) and convolutional neural network (CNN) models were built for maturity prediction. The results show that the x-axis region of interest selection method outperforms the other methods, and the SPA-BP model based on this method performs best. An accuracy of 99.19% for the correction set and 100% for the prediction set was achieved when only 0.03% of the wavelength was used. This groundbreaking study highlights the significant potential of hyperspectral technology for in situ assessment of citrus ripening stages. Furthermore, it offers crucial technical support and serves as a valuable reference for the advancement of precision agriculture.

Why it matches plant phenotyping methods柑橘果実の成熟段階という植物器官の状態を、ハイパースペクトル画像、ROI選択、前処理、波長選択、機械学習により非破壊推定する方法が研究の中心である。

abstractutilized hyperspectral technology to collect data from 22 fruit trees in an orchard

XFruitSeg—A general plant fruit segmentation model based on CT imaging →

Identification of the phenotypes of fruits is critical for understanding complex genetic traits. Computed tomography (CT) imaging technology enables the noninvasive acquisition of three-dimensional images of fruit interiors, thus providing a robust data foundation for phenotypic analysis. Accurate segmentation of internal fruit tissues is essential, as it directly influences the accuracy and reliability of the results. Current methods are not optimized for the unique features of plant fruit images. This study introduces XFruitSeg, which is a general deep learning model for segmenting plant fruit CT images. The model uses a U-shaped encoder–decoder architecture and integrates multitask learning. A large convolutional kernel network, RepLKNet, expands the receptive field for feature extraction. Multiscale skip connections and a deep supervision mechanism improve the model's capacity to learn features of various sizes, and a contour feature learning branch specifically targets the interorganizational boundaries. An optimized composite loss function enhances the model's robustness when applied to imbalanced categories. Additionally, a dataset named XrayFruitData was established, which contains high-resolution images of twelve plant fruit varieties, with accurate annotations for orange, mangosteen, and durian fruits for model evaluation. Compared with four mainstream advanced models, XFruitSeg achieved superior segmentation performance on the orange, mangosteen, and durian datasets, with mean Dice coefficients of 95.21 ​%, 93.24 ​%, and 94.70 ​% and mean intersection over union (mIoU) scores of 91.09 ​%, 87.91 ​%, and 90.35 ​%, respectively. The results of extensive ablation experiments demonstrate the effectiveness of each component. Therefore, the proposed XFruitSeg model has been proven to be beneficial for high-precision analysis of internal fruit phenotyping traits.

Why it matches plant phenotyping methods果実CT画像から内部組織を抽出するセグメンテーション手法を開発し、データセット構築と性能比較・検証を行っており、植物表現型取得の方法が中心である。

abstractThis study introduces XFruitSeg, which is a general deep learning model for segmenting plant fruit CT images.

Detection of soluble solid content in citrus fruit using near-infrared spectroscopy with machine learning regression: An exploration of the influence of sampling positions →

Near-infrared (NIR) spectroscopy has been widely used as the non-destructive technique for fruit SSC measurement. This study explored the combination of NIR spectroscopy and machine learning regression to predict SSC in 288 citrus fruits (cv. Ponkan mandarin) considering the influence of sampling positions. This research analyzed spectral variations in different sampling positions, as well as the average spectra. Machine learning algorithms, including support vector regression (SVR) and partial least squares regression (PLSR), were used to establish the prediction models for SSC using the single-position spectra, spectra of all sampling positions and the average spectra. Feature wavelengths were identified by the combination of correlation analysis and regression coefficient of PLSR models from the single-position spectra and the average spectra. Using the full spectra or feature wavelengths, the models based on the average spectra significantly outperformed those based on the sample-position spectra, indicating that the average spectra may be more suitable for SSC prediction of Ponkan mandarin. This study indicated the variations among different sampling positions and samples were one of the key factors affecting the precise and robust models for SSC prediction, and future attempts should be conducted to cover the sample variations improve the model robustness and generalization ability.

Why it matches plant phenotyping methods柑橘果実の可溶性固形分という植物器官形質を、NIR分光と機械学習で非破壊推定する手法の構築・比較が研究の中心であり、サンプリング位置とモデル性能も評価している。

abstractNear-infrared (NIR) spectroscopy has been widely used as the non-destructive technique for fruit SSC measurement.

Early detection of Citrus Huanglongbing by UAV remote sensing based on MGA-UNet. →

Citrus Huanglongbing (HLB), also known as citrus greening, is a severe disease that has caused substantial economic damage to the global citrus industry. Early detection is challenging due to the lack of distinctive early symptoms, making current diagnostic methods often ineffective. Therefore, there is an urgent need for an intelligent and timely detection system for HLB. This study leverages multispectral imagery acquired via unmanned aerial vehicles (UAVs) and deep convolutional neural networks. This study introduce a novel model, MGA-UNet, specifically designed for HLB recognition. This image segmentation model enhances feature transmission by integrating channel attention and spatial attention within the skip connections. Furthermore, this study evaluate the comparative effectiveness of high-resolution and multispectral images in HLB detection, finding that multispectral imagery offers superior performance. To address data imbalance and augment the dataset, this study employ a generative model, DCGAN, for data augmentation, significantly boosting the model's recognition accuracy. Our proposed model achieved a mIoU of 0.89, a mPA of 0.94, a precision of 0.95, and a recall of 0.94 in identifying diseased trees. The intelligent monitoring method for HLB presented in this study offers a cost-effective and highly accurate solution, holding considerable promise for the early warning of this disease.

Why it matches plant phenotyping methodsUAVマルチスペクトル画像と新規セグメンテーションモデルにより、感染樹という植物の病害状態を直接推定する手法を開発・評価しており、フェノタイピング手法が中心である。

abstractThis study leverages multispectral imagery acquired via unmanned aerial vehicles (UAVs) and deep convolutional neural networks.

Classifying Storage Temperature for Mandarin ( Citrus reticulata L.) Using Bioimpedance and Diameter Measurements with Machine Learning. →

Mandarin ( Citrus reticulata L.) is consumed worldwide. Improper storage temperatures cause flavor loss and shorten shelf lives, reducing marketability. Mandarins' quality is difficult to assess visually, as they show no apparent changes during storage. Therefore, a simple, non-destructive method is needed to assess their freshness as affected by temperature. This work utilized non-invasive bioimpedance spectroscopy (BIS) on mandarins stored at different temperatures. Eight machine learning (ML) models were trained with bioimpedance data to classify storage temperature. Also, we confirmed whether integrating diameter and time-series changes into the bioimpedance could improve the ML models' accuracies by minimizing sample variations. Additionally, we evaluated the effectiveness of equivalent circuit (EC) parameters derived from bioimpedance data for ML training. Although slightly less accurate than using raw bioimpedance data, EC parameters can efficiently reduce data dimensionality. Among all models, the SVM model trained with changes in bioimpedance integrated with diameter data achieved the highest accuracy of 0.92. It was a significant improvement compared to the accuracy of 0.76 achieved when using only the raw bioimpedance data. Thus, this study suggests a novel method of integrating diameter and bioimpedance changes to assess the storage temperature of mandarins. This approach can also be applied to other fruits when utilizing BIS.

Why it matches plant phenotyping methodsマンダリン果実の保存温度・鮮度状態を、バイオインピーダンスと径の非破壊測定および機械学習で推定する手法の開発・評価が中心である。

abstractThis work utilized non-invasive bioimpedance spectroscopy (BIS) on mandarins stored at different temperatures.

Integrating advanced deep learning techniques for enhanced detection and classification of citrus leaf and fruit diseases. →

In this study, we evaluate the performance of four deep learning models, EfficientNetB0, ResNet50, DenseNet121, and InceptionV3, for the classification of citrus diseases from images. Extensive experiments were conducted on a dataset of 759 images distributed across 9 disease classes, including Black spot, Canker, Greening, Scab, Melanose, and healthy examples of fruits and leaves. Both InceptionV3 and DenseNet121 achieved a test accuracy of 99.12%, with a macro average F1-score of approximately 0.986 and a weighted average F1-score of 0.991, indicating exceptional performance in terms of precision and recall across the majority of the classes. ResNet50 and EfficientNetB0 attained test accuracies of 84.58% and 80.18%, respectively, reflecting moderate performance in comparison. These research results underscore the promise of modern convolutional neural networks for accurate and timely detection of citrus diseases, thereby providing effective tools for farmers and agricultural professionals to implement proactive disease management, reduce crop losses, and improve yield quality.

Why it matches plant phenotyping methods柑橘の葉・果実画像から病害状態を分類する深層学習手法を複数モデルで比較評価しており、植物病害表現型の取得・分類が研究の中心である。

abstractwe evaluate the performance of four deep learning models, EfficientNetB0, ResNet50, DenseNet121, and InceptionV3, for the classification of citrus diseases from images.

Multi-scale remote sensing for sustainable citrus farming: Predicting canopy nitrogen content using UAV-satellite data fusion →

• Canopy Nitrogen Content (CNC) was estimated using a combination of UAV and Sentinel-2 data. • Developed models that utilize vegetation indices and tree structural data for CNC estimation. • Created spatial and temporal heat maps to visualize nitrogen content distribution in citrus trees. • Demonstrated a strong correlation between CNC and yield across three growing seasons. • Highlighted the potential for CNC applications to guide the development of site-specific nitrogen management strategies. Accurate monitoring of nitrogen (N) levels, while accounting for spatiotemporal variability is crucial for optimizing fertilization in citrus orchards. Traditional methods, such as frequent leaf and soil sampling followed by laboratory analysis, are costly, labor-intensive, and prone to human error. Remote sensing (RS) technologies, including unmanned aerial vehicles (UAVs) and satellite platforms, offer scalable and precise alternatives for N management. However, integrating these platforms poses challenges due to significant differences in spatial, temporal, and spectral resolution. This study presents a novel approach incorporating multispectral and temporal data from UAVs and Sentinel-2 satellites to estimate canopy N content (CNC) in citrus orchards. This method captures spatiotemporal variability across multiple citrus cultivars, aiming to enhance nitrogen use efficiency (NUE) while reducing environmental impact, ultimately promoting sustainable orchard management practices. The study was conducted in commercial citrus plots in the Hefer Valley, Israel, and spanned two phases. The first phase (May 2019 to April 2022) focused on four plots of the 'Newhall' cultivar, while the second phase expanded to twelve additional plots featuring five different citrus cultivars. The methodology consisted of six key steps: (1) Leaf samples from the study area were collected for laboratory nitrogen (N) analysis. (2) Acquiring and preprocessing bimonthly UAV multispectral images and Sentinel-2 satellite images to ensure data quality and consistency. (3) Segmenting individual trees using UAV imagery and extracting structural features through Structure-from-Motion (SfM) photogrammetry. (4) Processing images and extracting spectral and structural features relevant to N estimation. (5) Developing Random Forest (RF) models to estimate CNC using UAV-derived vegetation indices (VIs) and SfM data and combining these with Sentinel-2 VIs to generate canopy-scale CNC heatmaps. (6) Analyzing the relationship between CNC and yield to understand nitrogen dynamics and their impact on productivity. The integrated RF model, which combined UAV-VIs, Sentinel-2 VIs, and SfM-derived structural data, achieved superior performance (R² = 0.80, RMSE = 0.17 kg/m²) compared to models relying solely on UAV-VIs (R² = 0.68, RMSE = 0.23 kg/m²) or Sentinel-2 VIs (R² = 0.48, RMSE = 0.30 kg/m²). Additionally, CNC expressed as mass per tree demonstrated a strong positive correlation with yield (R² = 0.66), highlighting the relationship between nitrogen dynamics and orchard productivity. These results underscore the robustness of the integrated model and the clear advantage of multi-platform data fusion over single-source approaches. The study provides compelling evidence for the potential of combining UAV and Sentinel-2 data to improve CNC estimation and its correlation with yield in citrus orchards. The findings contribute to advancements in precision agriculture by offering a scalable, data-driven framework to enhance nutrient management and support sustainable orchard practices.

Why it matches plant phenotyping methodsUAV・衛星画像、SfM、特徴抽出、RFモデルを統合し、柑橘樹冠窒素含量という植物形質を推定する方法を開発・評価しており、表現型取得が研究の中心です。

abstractThis study presents a novel approach incorporating multispectral and temporal data from UAVs and Sentinel-2 satellites to estimate canopy N content (CNC) in citrus orchards.

Citrus Disease Detection Based on Dilated Reparam Feature Enhancement and Shared Parameter Head. →

Accurate citrus disease identification is essential for targeted orchard pesticide application. Current models struggle with accuracy and efficiency due to diverse leaf lesion patterns and complex orchard environments. This study presents YOLOv8n-DE, an improved lightweight YOLOv8-based model for enhanced citrus disease detection. It introduces the DR module structure for effective feature enhancement and the Detect_Shared architecture for parameter efficiency. Evaluated on public and orchard-collected datasets, YOLOv8n-DE achieves 97.6% classification accuracy, 91.8% recall, and 97.3% mAP, with a 90.4% mAP for challenging diseases. Compared to the original YOLOv8, it reduces parameters by 48.17%, computational load by 59.26%, and model size by 41.94%, while significantly decreasing classification and regression errors, and false positives/negatives. YOLOv8n-DE offers outstanding performance and lightweight advantages for citrus disease detection, supporting precision agriculture development in orchards.

Why it matches plant phenotyping methods柑橘葉の病斑に基づく植物病害状態の画像検出モデルを開発し、複数データセットで性能評価しており、フェノタイピング手法が中心である。

abstractThis study presents YOLOv8n-DE, an improved lightweight YOLOv8-based model for enhanced citrus disease detection.

Linking Citrus Fruit Cracking Intensity to Plant Water Status: Insights from UAV-Derived Metrics Validated by Ground-Based Data →

Citrus fruit cracking, a physical failure of the peel, causes yield losses of 10% to 35%, peaking during October-November. Water status of the tree and water flow into the fruit influence this phenomenon. with excessive irrigation during critical fruit development stages exacerbates cracking. As part of the EU-Horizon CrackSense project, this study is aimed to link citrus tree plant water status (PWS) to fruit cracking, emphasizing how deficit irrigation can reduce yield loss due to cracking. Using UAV and eco-physiological measurements, we developed models to predict PWS and its relationship with cracking intensity early in the season. The study, conducted in 2023-2024 in a commercial orchard near Kfar Chabad, Israel, tested four irrigation treatments: control, defined as the standard irrigation, two deficits irrigations regimes (50% of control) early and late in the season, and excessive irrigation (150% of control) throughout the season. Ground-based measurements included fruit and trunk diameter, stem water potential (SWP), stomatal conductance, plant area index (PAI), and growth rate (TG). UAV flights integrated multispectral, thermal, and LiDAR sensors to capture spatial-temporal variability in PWS. Canopy metrics, such as height, volume, LiDAR-based PAI, and spectral and thermal indices, were incorporated into PWS models. Results revealed significant differences in TG, SWP, and stomatal conductance for 50% of early and late deficit irrigation treatments compared to other treatments. Random forest models demonstrated strong predictive performance for SWP (R² > 0.77) and TG (R² > 0.76). LiDAR-derived PA correlated highly with field optical measurements (R² = 0.92), yield (R² = 0.67), and cracked fruit percentages (R² > 0.50). This study underscores the importance of precise irrigation management in reducing fruit cracking. It highlights the potential of remote sensing systems for predicting cracking and managing water status at the tree level. The developed models equip farmers with tools to apply controlled water stress, minimizing cracking and improving yield.

Why it matches plant phenotyping methodsUAVのマルチスペクトル・熱・LiDAR計測から樹体の水分状態、成長、樹冠形質、裂果状態を推定するモデルを開発し、地上計測で検証しており、表現型取得手法が研究の中心である。

abstractUsing UAV and eco-physiological measurements, we developed models to predict PWS and its relationship with cracking intensity early in the season.

Rapid and accurate detection of huanglongbing in citrus by elasticity testing using a piezoelectric finger. →

Rapid and sensitive detection of citrus huanglongbing (HLB) is critical for the control of this devastating disease. In this study, we have evaluated using a piezoelectric finger (PEF) with a 0.4 mm probe to measure the elastic modulus of a leaf to detect HLB in four different species of citrus including grapefruit (GFT), pumelo (PUM), lemon (LEM), and Valencia orange (VAL). Diseased citrus leaves were harvested from trees testing positive for the presence of Candidatus Liberibacter asiaticus (Las), the causal agent of HLB, and included both symptomatic leaves, which were blotchy mottle or yellowing and asymptomatic leaves, which did not display outward symptoms. Healthy leaves were harvested from trees testing negative for Las. The results indicated that the PEF elastic modulus test exhibited an overall 94% sensitivity and 90% specificity against the Las status of the trees for all four citrus types combined. Comparative quantitative real-time polymerase chain reaction (qPCR) tests on the same leaves showed an overall 89% sensitivity and 100% specificity against the Las status of the trees. While a Cohen-Kappa coefficient of 0.81 was obtained between the PEF and qPCR predictions, suggesting a "strong" agreement between the PEF and qPCR tests, a more detailed examination indicated that PEF was more sensitive overall in detecting the Las positive trees than qPCR, particularly from asymptomatic leaves for which PEF was 96% sensitive versus 78% sensitive by qPCR, indicating the potential of using PEF for early detection of HLB.

Why it matches plant phenotyping methods圧電プローブで葉の弾性率を測定し、カンキツHLBの感染状態を推定する手法を評価・検証しており、植物フェノタイピング手法が研究の中心です。

abstractwe have evaluated using a piezoelectric finger (PEF) with a 0.4 mm probe to measure the elastic modulus of a leaf to detect HLB

Citrus fruit diameter estimation in the field using monocular camera →

Accurate and efficient measurement of citrus fruit size is essential for managing tree form and estimating yields. Conventional manual methods are reliable but highly labour-intensive, while existing machine vision solutions often require specialised setups (e.g., distance calibration or 3D sensors). In this study, a low-cost, monocular-based framework that uses mature and healthy leaves as natural reference objects was proposed, eliminating the need for manual markers or complex camera parameter calibration. By compiling an offline leaf-size distribution from multiple citrus varieties, this method automatically converts fruit pixels to real-world diameters using the largest near-frontal leaf in each image. Further, the work integrates the deformable convolution (DNCv2) and shuffle attention (SA) into a YOLOv8 detector to improve occlusion handling, ensuring robust detection even when fruits are partially obscured by foliage. Extensive validation on three different citrus cultivars shows that leaf-size variability contributes less than 3.2% relative error in diameter estimation, while the overall approach achieves 93.14% accuracy and R² = 0.76. Key contributions include: (1) a novel monocular technique leveraging inherent orchard elements (leaves) as references, (2) advanced detection modules to tackle partial occlusion, (3) cross-variety validation demonstrating consistent performance, and (4) a fast, user-friendly workflow suitable for real-world orchard applications. Future work will explore multi-frame or multi-view strategies to further refine diameter measurement under heavy occlusion.

Why it matches plant phenotyping methods柑橘果実の直径という植物器官形質を、単眼画像と葉を基準に推定する手法を開発・検証しており、表現型取得が研究の中心である。

abstracta low-cost, monocular-based framework that uses mature and healthy leaves as natural reference objects was proposed

Integrating Machine Learning and RAG-Based Chatbot for Mandarin Orange Disease Detection in Hilly Region of Nepal →

Abstract Citrus farming, particularly Mandarin orange cultivation, is a crucial economic activity in Nepal’s hilly regions. However, disease detection and management remain major challenges. This study presents an efficient method for identifying and controlling five key citrus diseases affecting the Nepali orange market: black spot, canker, Huanglongbing (HLB), leaf miner, and sooty mold. We employ the MobileNetV2 model for disease prediction and a One-Class SVM model for initial leaf classification. Additionally, we integrate a Llama-3.2-11b-vision RAG-based chatbot, which analyzes leaf images and provides real-time guidance on disease prevention and orchard management. A mobile application has been developed to integrate the chatbot with a user-friendly interface, making it accessible for farmers. Our approach achieves 95.6% accuracy in disease identification and 85.6% accuracy in orange leaf classification. With its intuitive mobile platform and AI-driven chatbot, this system has the potential to transform citrus farming in Nepal by enabling timely interventions and improved disease management.

Why it matches plant phenotyping methods葉画像から柑橘病害を識別する手法と、その精度評価を中心に扱うため、植物の病害状態を推定するフェノタイピング手法として収録対象。

abstractThis study presents an efficient method for identifying and controlling five key citrus diseases affecting the Nepali orange market

Supervised Hyperspectral Band Selection Using Texture Features for Classification of Citrus Leaf Diseases with YOLOv8. →

Citrus greening disease (HLB) and citrus canker cause financial losses in Florida citrus groves via smaller fruits, blemishes, premature fruit drop, and/or eventual tree death. Management of these two diseases requires early detection and distinction from other leaf defects and infections. Automated leaf inspection with hyperspectral imagery (HSI) is tested in this study. Citrus leaves bearing visible symptoms of HLB, canker, scab, melanose, greasy spot, zinc deficiency, and a control class were collected, and images were taken with a line-scan HSI camera. YOLOv8 was trained to classify multispectral images from this image dataset, created by selecting bands with a novel variance-based method. The 'small' network using an intensity-based band combination yielded an overall weighted F1 score of 0.8959, classifying HLB and canker with F1 scores of 0.788 and 0.941, respectively. The network size appeared to exert greater influence on performance than the HSI bands selected. These findings suggest that YOLOv8 relies more heavily on intensity differences than on the texture properties of citrus leaves and is less sensitive to the choice of wavelengths than traditional machine vision classifiers.

Why it matches plant phenotyping methods柑橘葉の病徴をハイパースペクトル画像で取得し、バンド選択法とYOLOv8による分類性能を評価しており、植物病害状態の表現型取得・推定手法が研究の中心である。

abstractAutomated leaf inspection with hyperspectral imagery (HSI) is tested in this study.

Hyperspectral Imaging and Machine Learning for Huanglongbing Detection on Leaf-Symptoms. →

Huanglongbing is one of the most destructive diseases of citrus worldwide. Infected trees die due to the absence of practical cures. Thus, the removal of HLB-infected trees is one of the principal HLB managements for the regulation of disease spread. Here, we propose a non-destructive HLB detection method based on hyperspectral leaf reflectance. In total, 72 hyperspectral leaf images were collected in an HLB-invaded citrus orchard in Thailand and each image was visually distinguished into either any HLB symptom appearance (symptomatic) or no symptoms (asymptomatic) on the leaf. Principal component analysis was applied on the hyperspectral data and revealed 16 key wavelengths at red-edge to near-infrared regions (715, 718, 721, 724, 727, 730, 733, 736, 930, 933, 936, 939, 942, 945, 957, and 997 nm) that were characteristically differentiated in the symptomatic group. Seven models learnt on the spectral data at these 16 wavelengths were examined for the potential to separate these two image groups: random forest, decision tree, support vector machine, k-nearest neighbor, gradient boosting, logistic regression, linear discriminant. F1-score was employed to select the best-fit model to distinguish the two categories: random forest achieved the best score of 99.8%, followed by decision tree and k-nearest neighbor. The reliability of the visual grouping was evaluated by nearest neighbor matching and permutation test. These three models separated the two image categories as precisely as PCR results, indicating their potential as alternative tool instead of PCR.

Why it matches plant phenotyping methods葉のハイパースペクトル画像からHLB症状を非破壊的に検出・分類する手法を提案し、複数の機械学習モデルを比較・評価しているため、植物病徴の取得・推定が研究の中心です。

abstractHere, we propose a non-destructive HLB detection method based on hyperspectral leaf reflectance.

Citrus huanglongbing detection: A hyperspectral data-driven model integrating feature band selection with machine learning algorithms →

This study explored rapid detection techniques for citrus Huanglongbing (HLB), a disease that severely impacts global citrus production. The method based on hyperspectral technology combined with machine learning algorithms provides new ideas for rapid HLB identification. Algorithm selection is crucial for processing efficiency and hyperspectral data interpretation. Hyperspectral data from healthy, mild HLB-infected, and macular (not related to HLB) citrus leaves were captured using a hyperspectrometer, with qPCR validation. Three preprocessing methods were selected to preprocess the spectral data. Competitive Adaptive Reweighted Sampling (CARS) and Successive Projections Algorithm (SPA) were used to extract feature bands from the hyperspectral data, and the range of the number of filtered feature bands as a percentage of the full band was 22.87%–28.31% and 3.27%–4.17%, respectively. Five distinct algorithms were then employed to construct classification models. Upon evaluation, the SPA-STD-SVM algorithm combination proved most effective, boasting a 97.46% accuracy and a 98.55% recall rate. The results demonstrate that suitable machine learning algorithms can effectively classify the hyperspectral data of citrus leaves in three different states: healthy, mild HLB-infected, and macular. This provides an effective approach for using hyperspectral data to differentiate citrus Huanglongbing.

Why it matches plant phenotyping methods柑橘葉のハイパースペクトル取得と機械学習によるHLB感染状態の分類手法を開発・評価しており、植物病害状態の表現型取得が研究の中心である。

abstractThe method based on hyperspectral technology combined with machine learning algorithms provides new ideas for rapid HLB identification.

Machine learning driven benchtop Vis/NIR spectroscopy for online detection of hybrid citrus quality →

The aim of this study was to explore application of visible and near-infrared (Vis/NIR) spectroscopy combined with machine learning models for SSC and TA prediction of hybrid citrus. The Vis/NIR spectra of samples including navel-region, equator-region and multi-region combination spectra in navel-region and equator-region were collected using a benchtop equipment. The performance of SSC and TA prediction models with different region spectra, including partial least squares (PLS), random forest (RF), k-nearest neighbors (KNN), support vector machine (SVM) and multilayer feedforward neural network (MFNN), was assessed. The accuracy of SSC and TA prediction models with multi-region combination (raw) spectra was better compared to navel-region and equator-region, suggesting that multi-region combination spectra collection method was more suitable. Subsequently, the spectral pre-processing, including Savitzky-Golay smoothing (SGS), maximum normalization (MN), multiplicative scatter correction (MSC), linear baseline correction (LBC) and first derivative (1stD), were performed. The performance of SSC and TA prediction models with different pre-processing spectra was further compared. The PLS with SGS spectra (SGS-PLS) and MFNN with raw spectra (Raw-MFNN) exhibited superior validation effects for SSC and TA prediction, respectively. In a subsequent prediction in new samples, SGS-PLS achieved an RP² of 0.875, an RMSEP of 0.572% and a MAEP of 0.469% for SSC prediction, and Raw-MFNN achieved an RP² of 0.800, an RMSEP of 0.0322% and a MAEP of 0.0249% for TA prediction, indicating excellent generalization ability. These results indicate the great potential of benchtop Vis/NIR spectroscopy for online detection of hybrid citrus quality at mass-scale level.

Why it matches plant phenotyping methods柑橘果実のSSC・TAという植物器官形質を対象に、Vis/NIR分光と機械学習による推定法を比較・検証しており、形質取得手法が中心的である。

abstractexplore application of visible and near-infrared (Vis/NIR) spectroscopy combined with machine learning models for SSC and TA prediction of hybrid citrus

Machine learning driven benchtop Vis/NIR spectroscopy for online detection of hybrid citrus quality. →

The aim of this study was to explore application of visible and near-infrared (Vis/NIR) spectroscopy combined with machine learning models for SSC and TA prediction of hybrid citrus. The Vis/NIR spectra of samples including navel-region, equator-region and multi-region combination spectra in navel-region and equator-region were collected using a benchtop equipment. The performance of SSC and TA prediction models with different region spectra, including partial least squares (PLS), random forest (RF), k-nearest neighbors (KNN), support vector machine (SVM) and multilayer feedforward neural network (MFNN), was assessed. The accuracy of SSC and TA prediction models with multi-region combination (raw) spectra was better compared to navel-region and equator-region, suggesting that multi-region combination spectra collection method was more suitable. Subsequently, the spectral pre-processing, including Savitzky-Golay smoothing (SGS), maximum normalization (MN), multiplicative scatter correction (MSC), linear baseline correction (LBC) and first derivative (1stD), were performed. The performance of SSC and TA prediction models with different pre-processing spectra was further compared. The PLS with SGS spectra (SGS-PLS) and MFNN with raw spectra (Raw-MFNN) exhibited superior validation effects for SSC and TA prediction, respectively. In a subsequent prediction in new samples, SGS-PLS achieved an R P 2 of 0.875, an RMSEP of 0.572% and a MAEP of 0.469% for SSC prediction, and Raw-MFNN achieved an R P 2 of 0.800, an RMSEP of 0.0322% and a MAEP of 0.0249% for TA prediction, indicating excellent generalization ability. These results indicate the great potential of benchtop Vis/NIR spectroscopy for online detection of hybrid citrus quality at mass-scale level.

Why it matches plant phenotyping methods柑橘のSSC・TAという果実形質を、Vis/NIR分光と機械学習で非破壊推定する測定・解析手法が研究の中心であり、複数の前処理・モデル比較と新規試料での検証も行っている。

abstractexplore application of visible and near-infrared (Vis/NIR) spectroscopy combined with machine learning models for SSC and TA prediction of hybrid citrus

CitrusGAN: sparse-view X-ray CT reconstruction for citrus based on generative adversarial networks →

3D phenotyping of the external and internal structures is important to breed new fruit species. As manual phenotyping is error-prone and time-consuming, developing high-throughput solutions with enhanced precision and low costs is necessary. This study presents CitrusGAN, a generative adversarial network-based method to reconstruct 3D citrus CT models from sparse-view X-ray images. The input X-rays are arranged in orthogonal pairs to provide additional information, and customized loss functions enable more effective learning of the mapping from 2D X-ray features to 3D CT volumes. Experimental results show that 6 views can generate high-quality citrus CT volumes, with a structural similarity index of 92.1% and a peak signal-to-noise ratio of 26.374 dB compared with the real CT models. Moreover, the morphology of the generated model can be conveniently measured in the 3D space, facilitating the extraction of phenotypic traits including fruit length, width, height, volume, surface area, peel thickness, number of segments, and edible rate with high precision. As X-rays can be obtained using low-cost X-ray machines with high efficiency, the proposed method can be potentially developed into high-throughput equipment for fruit production lines or portable devices to realize in-field phenotyping.

Why it matches plant phenotyping methods柑橘の疎視野X線から3D CTモデルを再構成し、形態形質を抽出する手法を開発・評価しており、フェノタイピング手法が研究の中心である。

abstractThis study presents CitrusGAN, a generative adversarial network-based method to reconstruct 3D citrus CT models from sparse-view X-ray images.

GACNet: A Geometric and Attribute Co-Evolutionary Network for Citrus Tree Height Extraction From UAV Photogrammetry-Derived Data →

The undulating terrain and complex backgrounds of citrus plantations introduce nonlinear variations that significantly impede the high-precision estimation of citrus tree heights from remote sensing data. To overcome these obstacles, we introduce a novel geometric and attribute co-evolutionary network, tailored for extracting citrus tree heights using unmanned aerial vehicle photogrammetry-derived data. Our approach integrates a multisource feature interaction module with a multisource feature aggregation module, fostering the co-evolution of deep feature responses across various datasets. Notably, this includes a sophisticated triple-feature interaction mechanism that considers position, channel, and spatial correlation to enhance the aggregation of geometric features. In addition, we employ a multilevel feature aggregation decoder leveraging cross-attention, ensuring attribute context consistency and facilitating efficient tree height extraction. Quantitative analysis across datasets reveals our method's superior performance, with a 2% –7% increase in mean intersection over union for canopy segmentation and a robust correlation of 0.77 between estimated and reference tree heights, accompanied by an MAE of 0.25 m and an RMSE of 0.38 m. Comparative experiments indicate that our method outperforms current state-of-the-art networks, showing resilience to terrain undulations and offering reliable cross-region and cross-scale tree height estimation capabilities.

Why it matches plant phenotyping methodsUAV写真測量データから柑橘樹の樹高を抽出する新規ネットワークを開発・評価しており、植物形質の取得手法が研究の中心である。

abstractwe introduce a novel geometric and attribute co-evolutionary network, tailored for extracting citrus tree heights using unmanned aerial vehicle photogrammetry-derived data.

Multiple light sources excited fluorescence image-based non-destructive method for citrus Huanglongbing disease detection →

Citrus Huanglongbing (HLB) poses a significant threat to citrus orchards. Timely HLB screening of citrus trees is essential for citrus orchard management. This study developed a multi-excitation fluorescence imaging system based on the correlation between HLB stress and flavonoid fluorescence characteristics. The feasibility of using fluorescent images to classify healthy, macular (nutrient-deficient, not relate to HLB) and HLB-infected citrus fruits was explored. Initially, three-dimensional fluorescence spectra of citrus peels at maturity were scanned and the obtained Excitation-Emission Matrices (EEMs) were analyzed to screen four fluorescence characteristic regions (FCR1-FCR4) that were sensitive to HLB-infected citrus. Subsequently, four fluorescence imaging conditions (G1: EX = 365 ± 20 nm, EM = 525 ± 20 nm, G2: EX = 415 ± 20 nm, EM = 525 ± 20 nm, B1: EX = 308 ± 20 nm, EM = 450 ± 20 nm, and B2: EX = 365 ± 20 nm, EM = 450 ± 20 nm) were designed based on characteristic fluorescence bands. The imaging system primarily utilizes standard CMOS cameras and optical filters for image acquisition, offering significant advantages in terms of operational simplicity and cost-effectiveness. An HLB classification model was constructed using the Random Forest (RF) algorithm based on color feature parameters of fluorescence images, with a classification accuracy of up to 87.5 %. When the top 10 image feature parameters with the highest contribution rate were selected to construct the classification model considering the equipment cost, the accuracy is 83.33 %. This study demonstrated that fluorescence imaging utilizing flavonoid fluorescence characteristics enables non-destructive and rapid detection of citrus HLB. This approach provides valuable data and technical support for decision-making on spring orchard cleanup and control of HLB.

Why it matches plant phenotyping methods柑橘HLB感染状態を蛍光画像から推定する撮像システムと分類手法を開発・評価しており、植物状態の取得・抽出が研究の中心である。

abstractThis study developed a multi-excitation fluorescence imaging system based on the correlation between HLB stress and flavonoid fluorescence characteristics.

Citrus pose estimation under complex orchard environment for robotic harvesting →

The growth poses of citrus on trees are diverse. To ensure minimal loss during citrus harvesting, accurately estimating the pose of citrus is particularly important. To solve this problem, this research developed a real-time citrus pose estimation system based on neural networks and point cloud processing algorithms. Specifically, this method uses neural networks to identify citrus. After constructing the citrus point cloud, it is input into the Random Sample Consensus with Levenberg-Marquardt (RANSAC-LM) point cloud processing algorithm to obtain the citrus coordinates. Combined with citrus growth information, the pose is output. By analyzing the distribution of citrus poses, citrus poses convenient for end- effector harvesting are defined. To enhance the camera's ability to obtain information about citrus, a camera observation model is constructed to dynamically adjust the camera position. Through experiments, the appropriate deep learning target detection framework YOLO V5 is selected for citrus object detection. The precision (P), recall rate (R), and mean average precision (mAP) are 92.3 %, 79.1 %, and 88.5 % respectively. This network can handle detection tasks in real orchard environments. The original Random Sample Consensus (RANSAC) is improved by using the Levenberg-Marquardt (LM) nonlinear optimization method. Experimental results show that RANSAC-LM reduces the citrus center coordinate precision error from (0.2, 0.2, 2.3) mm to (0.1, 0.2, 1.4) mm, reduces the accuracy Spherical Error Probable (SEP) from 2.77 to 1.61, and finally reduces the citrus pose error from 5.72° to 2.43°. The efficiency of the proposed citrus pose estimation algorithm is 0.24 s. Deployed on a citrus picking robot, it verifies the feasibility of the algorithm and provides a new solution for the pose estimation problem of citrus harvesting robots.

Why it matches plant phenotyping methods柑橘果実の座標・生育姿勢という器官形質を点群処理とニューラルネットワークで推定し、精度を実験検証している。収穫対象の単なる検出を超えた姿勢計測法が中心である。

abstractthis research developed a real-time citrus pose estimation system based on neural networks and point cloud processing algorithms

Residual network-based feature extraction for automatic crop disease detection system using drone image dataset →

Purpose Diagnosing the crop diseases by farmers accurately with the naked eye can be challenging. Timely identification and treating these diseases is crucial to prevent complete destruction of the crops. To overcome these challenges, in this work a light-weight automatic crop disease detection system has been developed, which uses novel combination of residual network (ResNet)-based feature extractor and machine learning algorithm based classifier over a real-time crop dataset. Design/methodology/approach The proposed system is divided into four phases: image acquisition and preprocessing, data augmentation, feature extraction and classification. In the first phase, data have been collected using a drone in real time, and preprocessing has been performed to improve the images. In the second phase, four data augmentation techniques have been applied to increase the size of the real-time dataset. In the third phase, feature extraction has been done using two deep convolutional neural network (DCNN)-based models, individually, ResNet49 and ResNet41. In the last phase, four machine learning classifiers random forest (RF), support vector machine (SVM), logistic regression (LR) and eXtreme gradient boosting (XGBoost) have been employed, one by one. Findings These proposed systems have been trained and tested using our own real-time dataset that consists of healthy and unhealthy leaves for six crops such as corn, grapes, okara, mango, plum and lemon. The proposed combination of Resnet49-SVM and ResNet41-SVM has achieved accuracy of 99 and 97%, respectively, for the images that have been collected from the city of Kurukshetra, India. Originality/value The proposed system makes novel contribution by using a newly proposed real time dataset that has been collected with the help of a drone. The collected image data has been augmented using scaling, rotation, flipping and brightness techniques. The work uses a novel combination of machine learning methods based classification with ResNet49 and ResNet41 based feature extraction.

Why it matches plant phenotyping methodsドローン画像から植物の健康状態・病害を推定する画像解析システムの開発が研究の中心であり、特徴抽出、分類、データセット構築と性能評価を含むため。

abstracta light-weight automatic crop disease detection system has been developed, which uses novel combination of residual network (ResNet)-based feature extractor and machine learning algorithm based classifier over a real-time crop dataset.

Assessing Huanglongbing Severity and Canopy Parameters of the Huanglongbing-Affected Citrus in Texas Using Unmanned Aerial System-Based Remote Sensing and Machine Learning. →

Huanglongbing (HLB), also known as citrus greening disease, is a devastating disease of citrus. However, there is no known cure so far. Recently, under Section 24(c) of the Federal Insecticide, Fungicide, and Rodenticide Act (FIFRA), a special local need label was approved that allows the trunk injection of antimicrobials such as oxytetracycline (OTC) for HLB management in Florida. The objectives of this study were to use UAS-based remote sensing to assess the effectiveness of OTC on the HLB-affected citrus trees in Texas and to differentiate the levels of HLB severity and canopy health. We also leveraged UAS-based features, along with machine learning, for HLB severity classification. The results show that UAS-based vegetation indices (VIs) were not sufficiently able to differentiate the effects of OTC treatments of HLB-affected citrus in Texas. Yet, several UAS-based features were able to determine the severity levels of HLB and canopy parameters. Among several UAS-based features, the red-edge chlorophyll index (CI) was outstanding in distinguishing HLB severity levels and canopy color, while canopy cover (CC) was the best indicator in recognizing the different levels of canopy density. For HLB severity classification, a fusion of VIs and textural features (TFs) showed the highest accuracy for all models. Furthermore, random forest and eXtreme gradient boosting were promising algorithms in classifying the levels of HLB severity. Our results highlight the potential of using UAS-based features in assessing the severity of HLB-affected citrus.

Why it matches plant phenotyping methodsUASリモートセンシングと機械学習を用いて、感染樹のHLB重症度、樹冠色、樹冠密度などの植物状態を推定・分類する方法が研究の中心である。

abstractWe also leveraged UAS-based features, along with machine learning, for HLB severity classification.

Terahertz Spectroscopy in Assessing Temperature-Shock Effects on Citrus →

Rapid assessment of physiological status is a precondition for addressing biological stress in trees so that they may recover. Environmental stress can cause water deficit in plants, while terahertz (THz) spectroscopy is sensitive to changes in aqueous solutions within organisms. This has given the THz sensor a competitive edge for evaluating plant phenotypes, especially under similar environmental stress, if there are existing differences in the corresponding THz information. In this study, we utilized THz technology in association with traditional weighing methods to explore physiological changes in citrus leaves under different temperature, duration, and stress treatment conditions. It was found that the higher the temperature and the longer the exposure duration, the more severe the reduction in the relative absorption coefficient. There was a positive correlation between the trends and the increase in the ion permeability of cells. In addition, based on the effective medium theory, THz spectral information can be transformed into information on free water and bound water in the leaves. Under different treatment conditions, water content shows different trends and degrees of change on the time scale, and accuracy was verified by traditional weighing methods. These findings revealed that characteristics of THz information can serve as a simple and clear indicator for judging a plant's physiological status.

Why it matches plant phenotyping methodsTHz分光法で柑橘葉の生理状態・水分状態を推定し、従来の秤量法で精度検証しており、植物表現型取得法が中心である。

abstractwe utilized THz technology in association with traditional weighing methods to explore physiological changes in citrus leaves

Candidatus Liberibacter asiaticus infection alters the reflectance profile in asymptomatic citrus plants. →

Background Huanglongbing (HLB) is the primary and most destructive disease affecting citrus, caused by a pathogen transmitted by an insect vector, Diaphorina citri. There are no curative methods for the disease, and rapid and accurate methods are needed for early detection in the field, even before symptoms appear. These will facilitate the faster removal of infected trees, preventing the spread of the bacteria through commercial citrus orchards. Results It was possible to determine ranges of hyperspectral bands that demonstrated significant differences in relative reflectance between treatments consisting of healthy and infected plants from the first days of evaluation, when plants infected with 'Candidatus Liberibacter asiaticus' (CLas) were still in the asymptomatic stage of the disease. From the Week 2 of evaluation [58 days after infection (DAI) of plants] until the last week, spectral differences were detected in the red edge region (660-750 nm). From the Week 6 onwards (86 DAI), spectral differences between healthy and symptomatic plants were observed in bands close to the visible region (520-680 nm). Conclusion Spectral differences were detected in the leaves of C. sinensis infected by CLas before the appearance of symptoms, making it feasible to use the hyperspectral sensor to monitor the disease. Our results indicate the need for future studies to validate the use of hyperspectral sensors for managing and detecting HLB in commercial citrus orchards, contributing to the integrated management of the disease. © 2024 Society of Chemical Industry.

Why it matches plant phenotyping methods柑橘葉のハイパースペクトル反射を用いて、無症状段階の感染植物の病態を検出・モニタリングする測定法の適用と技術的評価が中心であり、植物病害表現型の取得に該当する。

abstractrapid and accurate methods are needed for early detection in the field, even before symptoms appear.

Inline detection of citrus rind micro-wounds using contrast-enhanced X-ray imaging: A feasibility study →

Decay management is crucial in the citrus industry due to the rapid spread of infections through wounds. Despite the urgency, effective methodologies for screening citrus rind micro-wounds are lacking. This study presents a preliminary investigation into real-time detection of citrus rind micro-wounds using contrast-enhanced X-ray imaging. This method highlights and magnifies rind wounds in X-ray images. The process involves immersing fruit in a contrast solution, capturing three sequential X-ray images, and then washing off residual contrast. Satsuma mandarins were used, with potassium iodide (KI) as the contrast agent duo to its distinct contrast properties on rind wounds coupled with high safety levels. A Convolutional Neural Network-Long Short-Term Memory (CNN-LSTM) model with multi-head attention mechanisms was developed, achieving a detection accuracy of 97.19 %. Post-radiography assessments showed minimal effects on the fruit's external appearance and internal quality, though a slight weight loss was observed. These results demonstrate the proposed method's effectiveness in detecting citrus rind micro-wounds, offering a promising approach for enhancing decay management in the citrus industry.

Why it matches plant phenotyping methods柑橘果皮の微小創傷という植物器官の状態を、造影X線画像とCNN-LSTMでリアルタイム検出する手法の開発が中心であり、植物病害・損傷状態のフェノタイピングに該当する。

abstractThis study presents a preliminary investigation into real-time detection of citrus rind micro-wounds using contrast-enhanced X-ray imaging.

Citrus yield estimation for individual trees integrating pruning intensity and image views →

Accurately estimating the yield of citrus fruit on individual trees is essential for precise orchard management and the income of producers. However, estimating the yield of citrus fruit from images of trees remains challenging among different processes of tree pruning and image acquisition. This study adopted a deep learning based detection model to count fruit in tree images and machine learning models to estimate the yield of individual trees from the fruit count. Trees under four levels of pruning intensity (no pruning, 0–5 %, 5–10 %, and 10–15 % of new sprouts pruned) and imaged from three different views (two, four, and six images per tree) to determine the optimal conditions for yield estimation. The variables considered for yield estimation included fruit count, pruning intensity and image views. Dataset containing 1200 tree images were used to train and test four machine learning models: random forest, support vector machine, extreme gradient boosting (XGBoost), and generalized linear model. The XGBoost model achieved the lowest errors in both training and testing. The optimal yield estimation occurs when there are two, four, and six image views and trees that have been pruned >10 %, 5–10 %, and ≤5 %, respectively. The findings can enhance the accuracy of image based citrus fruit yield estimation for individual trees and reveal the influences of pruning and image views.

Why it matches plant phenotyping methods個体樹の果実画像から果実数を検出し、機械学習で収量を推定する手法が研究の中心であり、画像枚数と剪定条件による精度も評価している。

abstractThis study adopted a deep learning based detection model to count fruit in tree images and machine learning models to estimate the yield of individual trees from the fruit count.

Channel randomisation: Self-supervised representation learning for reliable visual anomaly detection in speciality crops →

Modern, automated quality control systems for speciality crops utilise computer vision together with a machine learning paradigm exploiting large datasets for learning efficient crop assessment components. To model anomalous visuals, data augmentation methods are often developed as a simple yet powerful tool for manipulating readily available normal samples. State-of-the-art augmentation methods embed arbitrary “structural” peculiarities in normal images to build a classifier of these artefacts (i.e., pretext task), enabling self-supervised representation learning of visual signals for anomaly detection (i.e., downstream task). In this paper, however, we argue that learning such structure-sensitive representations may be suboptimal for agricultural anomalies (e.g., unhealthy crops) that could be better recognised by a different type of visual element like “colour”. To be specific, we propose Channel Randomisation (CH-Rand)—a novel data augmentation method that forces deep neural networks to learn effective encoding of “colour irregularities” under self-supervision whilst performing a pretext task to discriminate channel-randomised images. Extensive experiments are performed across various types of speciality crops (apples, strawberries, oranges, and bananas) to validate the informativeness of learnt representations in detecting anomalous instances. Our results demonstrate that CH-Rand’s representations are significantly more reliable and robust, outperforming state-of-the-art methods (e.g., CutPaste) that learn structural representations by over 43% in Area Under the Precision–Recall Curve (AUC–PR), particularly for strawberries. Additional experiments suggest that adopting the L∗a∗b∗ colour space and “curriculum” learning in the pretext task — gradually disregarding channel combinations for unrealistic outcomes — further improves downstream-task performance by 16% in AUC–PR. In particular, our experiments employ Riseholme-2021, a novel speciality crop dataset consisting of 3.5K real strawberry images gathered in situ from the real farm, along with the Fresh & Stale public dataset. All our code and datasets are made publicly available online to ensure reproducibility and encourage further research in agricultural technologies.

Why it matches plant phenotyping methods作物画像から異常・不健全状態を検出する画像解析手法を開発し、複数作物で検証しているため、植物状態の取得・推定が研究の中心である。

abstractwe propose Channel Randomisation (CH-Rand)—a novel data augmentation method that forces deep neural networks to learn effective encoding of “colour irregularities” under self-supervision

Image dataset: Optimizing growth of nonembryogenic citrus tissue cultures using response surface methodology. →

The data are images of Valencia sweet orange nonembryogenic tissue grown on different culture media that varied in the composition of the mineral nutrients from three experiments. Experiment 1 was a 5-factor d-optimal response surface design of five groupings of the component salts that make up Murashige and Skoog (MS) basal salt medium. Experiment 2 was a 3-factor d-optimal response surface design of extended ranges of factors 1, 2, and 3 from Experiment 1. Experiment 3 was thirteen formulations that were predicted using the prediction model generated from the 5-factor RSM from Experiment 1. The predictions were for two types of growth. One, points were predicted where growth was equal to MS medium (the standard), and two, points predicted with growth greater than MS medium by a minimum of 25%. An image representative of each formulation in each of the experiments makes up the dataset. The data will be useful for 1) visualizing the effects of the diverse mineral nutrient compositions, effects that may not be fully captured with single measure metrics; 2) development of image analysis applications via computer vision and segmentation algorithms for additional insight or for more rapid and possibly accurate assessment of tissue growth and quality; and 3) as an educational resource to learn how to use multifactor experimental designs to assess in vitro growth.

Why it matches plant phenotyping methods植物組織の成長画像データセットを提供し、画像解析・コンピュータビジョンによる成長および品質評価への利用を明示しており、表現型取得・抽出が中心的なデータ資源である。

abstractThe data are images of Valencia sweet orange nonembryogenic tissue grown on different culture media

Early detection of fungal infection in citrus using biospeckle imaging →

Fungi are among the leading defects causing severe economic losses in the citrus market. Early fungal infection diagnosis is crucial to avoid their propagation throughout production. This paper presents a new non-destructive and accurate method based on biospeckle imaging for the early identification of green mold due to Penicillium digitatum. First, the time and frequency domain properties of biospeckle signals of citrus inoculated with fungal suspension and sterile water were investigated. Three biospeckle parameter images were acquired to analyze changes in biospeckle activity during citrus infection. Next, five numerical parameters were extracted from two regions (injected and infected) of citrus to characterize the pattern of activity change in citrus from healthy to decay. Then, parameters were combined with support vector machine (SVM) and artificial neural network (ANN) classification methods to build fungal infection prediction models. Parameter sensitivity analysis was performed on the best-performing model. The results show that the early decay properties of the infected area appeared in biospeckle parameter images on the second day of infection, earlier than in RGB images. The five indexes in two citrus regions showed consistent variations in the biological activity of citrus at different infection stages. The parameters in the infected region could precede the appearance of visible fungal infection by one to three days. The ANN-based one-day-in-advance prediction model, two-days-in-advance prediction model and three-days-in-advance prediction model achieved 93.9%, 89.3%, and 86.4% discriminant accuracy in the prediction set, respectively. And the SVM-based one-day-in-advance prediction model, two-days-in-advance prediction model and three-days-in-advance prediction model achieved 90.2%, 83.9%, and 81.4% discriminant accuracy in the prediction set, respectively. Therefore, our results show that early fungal infections in citrus can be identified with biospeckle technology and discriminant analysis. Consequently, our techniques can be used in agriculture research to classify fruit fungal infections efficiently and effectively, developing biospeckle imaging technology use in the related sector.

Why it matches plant phenotyping methods柑橘果実の感染状態を対象に、バイオスペックル画像から特徴量を抽出し、SVM/ANNで早期感染を予測する手法を開発・評価しており、植物病害表現型の取得が中心である。

abstractThis paper presents a new non-destructive and accurate method based on biospeckle imaging for the early identification of green mold due to Penicillium digitatum.

Stem water content is crucial to support fruit tree functioning during heatwaves in a Mediterranean climate →

Droughts are expected to intensify in the Mediterranean region due to climate change, yet the effect of these highly variable events on local trees is unknown. To study the particular effect of heatwaves in orchards, where soil-drought can be mitigated by irrigation, we propose a heatwave definition that focuses on atmospheric stress and its consequences, by relating the intensity of high VPD events to losses in tree stem-water storage (StWS). We found that the sensitivity and resilience of StWS to heatwaves is species-specific, and varies among species with different water-management strategies (e.g., isohydric orange and anisohydric mango trees, p < 10 −3 ). Navel orange trees were sensitive to heatwaves starting at the 80th percentile of VPD in early spring, and once irrigation began, despite the harsh Mediterranean summer temperatures, StWC increased to 0.57 g cm −3 , slightly greater than the StWC of the earlier wet season (approximately 0.55 g cm −3 ). Oppositely, there was a net reduction in StWC in Shelly mango trees from 0.75 to 0.69 g cm −3 between the two seasons, as sensitivity to heatwaves increased from the 90th to the 80th percentile in spring and summer, respectively. By first quantifying heatwaves and relating this new variable to changes in StWS, we were able to describe the sensitivity of each species according to the rarity of the heatwave events by VPD percentile, and their resilience to heatwaves over seasons based on the corresponding net changes in StWC. Though the experiment in this study was performed in a Mediterranean climate, hotter-droughts are rising globally and the framework developed here for quantifying and measuring the effect of heatwaves can be broadly applied across geographic locations.

Why it matches plant phenotyping methods熱波をVPD percentileで定量化し、樹幹水分貯蔵・含水量の変化から樹木の生理的応答を評価する汎用的フレームワークが研究の中心であり、単なる routine 測定ではない。

abstractwe propose a heatwave definition that focuses on atmospheric stress and its consequences, by relating the intensity of high VPD events to losses in tree stem-water storage (StWS).

MSDD-YOLOX: An enhanced YOLOX for real-time surface defect detection of oranges by type →

Using an online high-throughput detection system for sorting oranges during the post-harvest process helped improve the commercialization level of the oranges industry. Surface defects on oranges created a poor first impression for consumers, making the rapid detection of oranges surface defects a primary concern for online sorting systems. However, due to variations in defect size and the visual similarity of different defects, there were still some challenges in detecting and identifying various surface defects on orange fruits based on their types. To address these challenges, this study first categorized surface defects on oranges into three major categories: deformity, scarring, and disease spot, based on their causes and potential post-harvest losses. Subsequently, to achieve real-time detection of orange surface defects on the orange sorting machine, a YOLOX-based real-time multi-type surface defect detection algorithm (MSDD-YOLOX) for oranges was proposed. This algorithm significantly improved the detection effectiveness of scarring at different scales by introducing neck network residual connections and cascading of the neck network. To address the issue of missed detections in texture-based defects and improve the regression of predicted bounding boxes, focal loss and Complete-IoU (CIoU) were employed in the algorithm. The results showed that MSDD-YOLOX achieved F1 values of 88.3 %, 80.4 %, and 92.7 % for the detection of deformity, scarring, and disease spot, respectively, with an overall detection F1 value of 90.8 %. These values represented improvements of 13.1 %, 10.2 %, 4.5 %, and 6.4 %, respectively, compared to the baseline model. Furthermore, compared to other deep learning object detectors, namely Faster RCNN, RetinaNet, FCOS, and Swin-Transformer, the proposed algorithm achieved optimal detection accuracy. Additionally, the MSDD-YOLOX model had a compact size of only 8.98 M, enabling real-time detection on the fruit grading line with an inference speed of up to 64.2FPS. Another innovation of this research was the external validation conducted on green oranges from Hainan and mandarins from Zhejiang. The results of external testing demonstrated that MSDD-YOLOX achieved overall F1 values of 90.6 % and 81.1 % for citrus fruits in these two regions, effectively proving the online deployment capability of MSDD-YOLOX and providing a robust solution for external defect detection in citrus fruits.

Why it matches plant phenotyping methodsオレンジ果実表面の形態・病斑状態を画像から抽出する手法を開発し、複数モデル比較と外部検証、オンライン実装性能評価まで行っており、表現型取得が研究の中心である。

abstracta YOLOX-based real-time multi-type surface defect detection algorithm (MSDD-YOLOX) for oranges was proposed

Machine learning-driven assessment of biochemical qualities in tomato and mandarin using RGB and hyperspectral sensors as nondestructive technologies. →

Estimation of fruit quality parameters are usually based on destructive techniques which are tedious, costly and unreliable when dealing with huge amounts of fruits. Alternatively, non-destructive techniques such as image processing and spectral reflectance would be useful in rapid detection of fruit quality parameters. This research study aimed to assess the potential of image processing, spectral reflectance indices (SRIs), and machine learning models such as decision tree (DT) and random forest (RF) to qualitatively estimate characteristics of mandarin and tomato fruits at different ripening stages. Quality parameters such as chlorophyll a (Chl a), chlorophyll b (Chl b), total soluble solids (TSS), titratable acidity (TA), TSS/TA, carotenoids (car), lycopene and firmness were measured. The results showed that Red-Blue-Green (RGB) indices and newly developed SRIs demonstrated high efficiency for quantifying different fruit properties. For example, the R2 of the relationships between all RGB indices (RGBI) and measured parameters varied between 0.62 and 0.96 for mandarin and varied between 0.29 and 0.90 for tomato. The RGBI such as visible atmospheric resistant index (VARI) and normalized red (Rn) presented the highest R2 = 0.96 with car of mandarin fruits. While excess red vegetation index (ExR) presented the highest R2 = 0.84 with car of tomato fruits. The SRIs such as RSI 710,600, and R730,650 showed the greatest R2 values with respect to Chl a (R2 = 0.80) for mandarin fruits while the GI had the greatest R2 with Chl a (R2 = 0.68) for tomato fruits. Combining RGB and SRIs with DT and RF models would be a robust strategy for estimating eight observed variables associated with reasonable accuracy. Regarding mandarin fruits, in the task of predicting Chl a, the DT-2HV model delivered exceptional results, registering an R2 of 0.993 with an RMSE of 0.149 for the training set, and an R2 of 0.991 with an RMSE of 0.114 for the validation set. As well as for tomato fruits, the DT-5HV model demonstrated exemplary performance in the Chl a prediction, achieving an R2 of 0.905 and an RMSE of 0.077 for the training dataset, and an R2 of 0.785 with an RMSE of 0.077 for the validation dataset. The overall outcomes showed that the RGB, newly SRIs as well as DT and RF based RGBI, and SRIs could be used to evaluate the measured parameters of mandarin and tomato fruits.

Why it matches plant phenotyping methodsRGB画像・ハイパースペクトル反射と機械学習により、果実の品質・生理形質を非破壊推定する方法が研究の中心であり、検証用データで性能評価も行っている。

abstractAlternatively, non-destructive techniques such as image processing and spectral reflectance would be useful in rapid detection of fruit quality parameters.

Improved FasterViT model for citrus disease diagnosis. →

Despite advances in deep learning for plant leaf disease recognition, accurately distinguishing morphological features under varying environmental conditions continues to pose significant challenges. Traditional deep learning models often fail to effectively merge local and global information, especially in small-scale datasets, impairing performance and elevating training costs. Focusing on citrus diseases, we propose an improved FasterViT Model, an advanced hybrid CNN-ViT framework that builds upon the FasterViT model. The proposed model seamlessly integrates CNN's rapid local learning capabilities with ViT's global information processing strength, thereby effectively extracting complex textures and morphological features from images. Cross-stage alternating Mixup and Cutout methods are strategically employed to enhance model robustness and generalization capabilities, particularly valuable for fast learning on small-scale datasets by simulating a more diverse training environment. Triplet Attention and AdaptiveAvgPool mechanisms are utilized to reduce training costs and optimize training performance. The proposed model is tested on both our specially constructed small-scale citrus disease dataset called in-field small dataset and the comprehensive PlantVillage dataset. The experimental results demonstrated that the model exhibits the capability of fast learning and adaptation to small sample training in plant disease detection tasks, and demonstrates the effectiveness of our improvement approach in improving model accuracy and reducing training costs. Additionally, its exemplary performance in transfer learning scenarios underscores its adaptability and broad applicability. This study not only highlights the efficacy of the improved FasterViT model in addressing the complexities of plant disease image recognition but also pioneers a new paradigm for developing efficient, scalable, and robust classification systems.

Why it matches plant phenotyping methods植物葉の病徴・形態特徴を画像から認識する深層学習モデルを開発・評価しており、病害状態の表現型推定が研究の中心である。

abstractwe propose an improved FasterViT Model, an advanced hybrid CNN-ViT framework

Developing a prediction method for physicochemical characteristics of Pontianak Siam orange (Citrus suhuiensis cv. Pontianak) based on combined reflectance-Fluorescence spectroscopy and artificial neural network →

The slightly sweet and acidic taste offered by Pontianak Siam oranges is influenced by the total soluble solids (TSS) and acidity in the fruit, in which, measuring these attributes is commonly performed using instruments that potentially damage the fruit's structure, thus, impractical for fresh fruit products. Moreover, the process of classifying the quality of fresh oranges has been based on physical appearance, leading to subjective results. Correspondingly, the objective of the study is to develop a prediction method for the physicochemical characteristics of Pontianak Siam oranges based on VIS-NIR-Fluorescence spectroscopy and an artificial neural network (ANN) model. The method is applicable to classify oranges based on physicochemical characteristics without damaging the fruit's structure. As a result, the best model for classifying the maturity level of Pontianak Siam oranges was obtained using a dataset with all feature combined spectra, attaining a training accuracy of 0.99 and testing accuracy of 1. The best model for predicting TSS was obtained using all feature combined spectra dataset, attaining R2 training = 0.89 and R2 testing = 0.91. The best model for predicting acidity was obtained using all feature reflectance spectra datasets, attaining R2 training = 0.96 and R2 testing = 0.97. The best model for predicting fruit firmness was obtained using all feature reflectance spectra dataset, attaining R2 training = 0.97, R2 testing = 0.89. Overall, the combination of Vis-NIR reflectance and fluorescence spectroscopy have the potential to be applied for non-destructive assessment of citrus quality in terms of visual classification and maturity parameters prediction.

Why it matches plant phenotyping methodsVIS-NIR反射・蛍光分光とANNを用いて、柑橘果実の成熟度、TSS、酸度、硬度を非破壊推定する手法の開発が研究の中心であり、植物形質の取得・推定方法を評価している。

abstractthe objective of the study is to develop a prediction method for the physicochemical characteristics of Pontianak Siam oranges based on VIS-NIR-Fluorescence spectroscopy and an artificial neural network (ANN) model.

Detection of jelly orange granulation disease using a dual-input Resnet-Transformer model (DresT) based on acoustic vibration images and a novel acoustic vibration device →

Granulation is a common internal disease in citrus fruits, and it is difficult to distinguish fruits with granulation disease from their appearance. In this study, a novel acoustic vibration device based on a micro-LDV, a microphone and a resonance speaker was employed to collect acoustic vibration response signals of "Aiyuan 38" jelly orange. The one-dimensional acoustic vibration response signal was converted into acoustic vibration images, and a double-input Resnet-Transformer network (DresT) was constructed for extracting deep features in acoustic vibration images for identifying jelly-orange granulation disease. Firstly, train Drest and Resnet50 models using acoustic vibration images and compare the performance of Drest with that of Resnet50 (based on CNN). Then PLS-DA and SVM models are trained using acoustic vibration image texture features or acoustic vibration spectral features, and the performance is compared with the DresT model. The results showed that the DresT model trained using acoustic vibration images can accurately identify jelly orange granulation disease with a detection accuracy of 99.31 %. The F₁ of the model is 99.5 %, the accuracy is 99.01 %, and the recall is 100 %.

Why it matches plant phenotyping methods柑橘果実の内部病害を音響振動画像と専用デバイス、深層学習モデルで推定する手法が研究の中心であり、植物の病害状態を直接評価している。

abstracta novel acoustic vibration device based on a micro-LDV, a microphone and a resonance speaker was employed to collect acoustic vibration response signals

Analysis of the Pomelo Peel Essential Oils at Different Storage Durations Using a Visible and Near-Infrared Spectroscopic on Intact Fruit. →

Pomelo fruit pulp mainly is consumed fresh and with very little processing, and its peels are discarded as biological waste, which can cause the environmental problems. The peels contain several bioactive chemical compounds, especially essential oils (EOs). The content of a specific EO is important for the extraction process in industry and in research units such as breeding research. The explanation of the biosynthesis pathway for EO generation and change was included. The chemical bond vibration affected the prediction of EO constituents was comprehensively explained by regression coefficient plots and x-loading plots. Visible and near-infrared spectroscopy (VIS/NIRS) is a prominent rapid technique used for fruit quality assessment. This research work was focused on evaluating the use of VIS/NIRS to predict the composition of EOs found in the peel of the pomelo fruit ( Citrus maxima (J. Burm.) Merr. cv Kao Nam Pueng) following storage. The composition of the peel oil was analyzed by gas chromatography-mass spectrometry (GC-MS) at storage durations of 0, 15, 30, 45, 60, 75, 90, 105 and 120 days (at 10 °C and 70% relative humidity). The relationship between the NIR spectral data and the major EO components found in the peel, including nootkatone, geranial, β-phellandrene and limonene, were established using the raw spectral data in conjunction with partial least squares (PLS) regression. Preprocessing of the raw spectra was performed using multiplicative scatter correction (MSC) or second derivative preprocessing. The PLS model of nootkatone with full MSC had the highest correlation coefficient between the predicted and reference values (r = 0.82), with a standard error of prediction (SEP) of 0.11% and bias of 0.01%, while the models of geranial, β-phellandrene and limonene provided too low r values of 0.75, 0.75 and 0.67, respectively. The nootkatone model is only appropriate for use in screening and some other approximate calibrations, though this is the first report of the use of NIR spectroscopy on intact fruit measurement for its peel EO constituents during cold storage.

Why it matches plant phenotyping methods無傷果実の可視・近赤外分光とPLS回帰により、果皮精油成分を非破壊推定する測定・検量モデルを評価しており、植物器官の化学的形質取得が中心である。

abstractThis research work was focused on evaluating the use of VIS/NIRS to predict the composition of EOs found in the peel of the pomelo fruit

LSD-YOLO: Enhanced YOLOv8n Algorithm for Efficient Detection of Lemon Surface Diseases. →

Lemon, as an important cash crop with rich nutritional value, holds significant cultivation importance and market demand worldwide. However, lemon diseases seriously impact the quality and yield of lemons, necessitating their early detection for effective control. This paper addresses this need by collecting a dataset of lemon diseases, consisting of 726 images captured under varying light levels, growth stages, shooting distances and disease conditions. Through cropping high-resolution images, the dataset is expanded to 2022 images, comprising 4441 healthy lemons and 718 diseased lemons, with approximately 1-6 targets per image. Then, we propose a novel model lemon surface disease YOLO (LSD-YOLO), which integrates Switchable Atrous Convolution (SAConv) and Convolutional Block Attention Module (CBAM), along with the design of C2f-SAC and the addition of a small-target detection layer to enhance the extraction of key features and the fusion of features at different scales. The experimental results demonstrate that the proposed LSD-YOLO achieves an accuracy of 90.62% on the collected datasets, with mAP@50-95 reaching 80.84%. Compared with the original YOLOv8n model, both mAP@50 and mAP@50-95 metrics are enhanced. Therefore, the LSD-YOLO model proposed in this study provides a more accurate recognition of healthy and diseased lemons, contributing effectively to solving the lemon disease detection problem.

Why it matches plant phenotyping methodsレモン表面の健全・罹病状態を画像から推定するデータセットとYOLO手法の開発・評価が研究の中心であり、植物病害状態の画像フェノタイピングに該当する。

abstractThis paper addresses this need by collecting a dataset of lemon diseases, consisting of 726 images captured under varying light levels, growth stages, shooting distances and disease conditions.

Hyperspectral Image Analysis and Machine Learning Techniques for Crop Disease Detection and Identification: A Review →

Originally, the use of hyperspectral images was for military applications, but their use has been extended to precision agriculture. In particular, they are used for activities related to crop classification or disease detection, combining these hyperspectral images with machine learning techniques and algorithms. The study of hyperspectral images has a wide range of wavelengths for observation. These wavelengths allow for monitoring agricultural crops such as cereals, oilseeds, vegetables, and fruits, and other applications. In the ranges of these wavelengths, crop conditions such as maturity index and nutrient status, or the early detection of some diseases that cause losses in crops, can be studied and diagnosed. Therefore, this article proposes a technical review of the main applications of hyperspectral images in agricultural crops and perspectives and challenges that combine artificial intelligence algorithms such as machine learning and deep learning in the classification and detection of diseases of crops such as cereals, oilseeds, fruits, and vegetables. A systematic review of the scientific literature was carried out using a 10-year observation window to determine the evolution of the integration of these technological tools that support sustainable agriculture; among the findings, information on the most documented crops is highlighted, among which are some cereals and citrus fruits due to their high demand and large cultivation areas, as well as information on the main fruits and vegetables that are integrating these technologies. Also, the main artificial intelligence algorithms that are being worked on are summarized and classified, as well as the wavelength ranges for the prediction, disease detection, and analysis of other tasks of physiological characteristics used for sustainable production. This review can be useful as a reference for future research, based mainly on detection, classification, and other tasks in agricultural crops and decision making, to implement the most appropriate artificial intelligence algorithms.

Why it matches plant phenotyping methods作物のハイパースペクトル画像による病害検出・生理特性推定と機械学習手法を体系的にレビューしており、植物フェノタイピング手法が中心です。

abstractthis article proposes a technical review of the main applications of hyperspectral images in agricultural crops and perspectives and challenges that combine artificial intelligence algorithms such as machine learning and deep learning in the classification and detection of diseases of crops

A Detection Algorithm for Citrus Huanglongbing Disease Based on an Improved YOLOv8n. →

Given the severe impact of Citrus Huanglongbing on orchard production, accurate detection of the disease is crucial in orchard management. In the natural environments, due to factors such as varying light intensities, mutual occlusion of citrus leaves, the extremely small size of Huanglongbing leaves, and the high similarity between Huanglongbing and other citrus diseases, there remains an issue of low detection accuracy when using existing mainstream object detection models for the detection of citrus Huanglongbing. To address this issue, we propose YOLO-EAF (You Only Look Once-Efficient Asymptotic Fusion), an improved model based on YOLOv8n. Firstly, the Efficient Multi-Scale Attention Module with cross-spatial learning (EMA) is integrated into the backbone feature extraction network to enhance the feature extraction and integration capabilities of the model. Secondly, the adaptive spatial feature fusion (ASFF) module is used to enhance the feature fusion ability of different levels of the model so as to improve the generalization ability of the model. Finally, the focal and efficient intersection over union (Focal-EIOU) is utilized as the loss function, which accelerates the convergence process of the model and improves the regression precision and robustness of the model. In order to verify the performance of the YOLO-EAF method, we tested it on the self-built citrus Huanglongbing image dataset. The experimental results showed that YOLO-EAF achieved an 8.4% higher precision than YOLOv8n on the self-built dataset, reaching 82.7%. The F1-score increased by 3.33% to 77.83%, and the mAP (0.5) increased by 3.3% to 84.7%. Through experimental comparisons, the YOLO-EAF model proposed in this paper offers a new technical route for the monitoring and management of Huanglongbing in smart orange orchards.

Why it matches plant phenotyping methods柑橘黄龙病の画像から病徴を検出する改良物体検出法を開発し、自作画像データセットで性能検証しているため、植物病害状態のフェノタイピング手法が中心である。

abstractwe propose YOLO-EAF (You Only Look Once-Efficient Asymptotic Fusion), an improved model based on YOLOv8n.

YOLOC-tiny: a generalized lightweight real-time detection model for multiripeness fruits of large non-green-ripe citrus in unstructured environments. →

This study addresses the challenges of low detection precision and limited generalization across various ripeness levels and varieties for large non-green-ripe citrus fruits in complex scenarios. We present a high-precision and lightweight model, YOLOC-tiny, built upon YOLOv7, which utilizes EfficientNet-B0 as the feature extraction backbone network. To augment sensing capabilities and improve detection accuracy, we embed a spatial and channel composite attention mechanism, the convolutional block attention module (CBAM), into the head's efficient aggregation network. Additionally, we introduce an adaptive and complete intersection over union regression loss function, designed by integrating the phenotypic features of large non-green-ripe citrus, to mitigate the impact of data noise and efficiently calculate detection loss. Finally, a layer-based adaptive magnitude pruning strategy is employed to further eliminate redundant connections and parameters in the model. Targeting three types of citrus widely planted in Sichuan Province-navel orange, Ehime Jelly orange, and Harumi tangerine-YOLOC-tiny achieves an impressive mean average precision (mAP) of 83.0%, surpassing most other state-of-the-art (SOTA) detectors in the same class. Compared with YOLOv7 and YOLOv8x, its mAP improved by 1.7% and 1.9%, respectively, with a parameter count of only 4.2M. In picking robot deployment applications, YOLOC-tiny attains an accuracy of 92.8% at a rate of 59 frames per second. This study provides a theoretical foundation and technical reference for upgrading and optimizing low-computing-power ground-based robots, such as those used for fruit picking and orchard inspection.

Why it matches plant phenotyping methods柑橘の熟度状態を画像から推定する軽量検出モデルを開発し、精度・速度を比較評価しており、植物状態の取得手法が中心である。

abstractWe present a high-precision and lightweight model, YOLOC-tiny, built upon YOLOv7

Delineating citrus management zones using spatial interpolation and UAV-based multispectral approaches →

The ability to delineate site-specific management zones is a key feature for precision agriculture applications. In this study, a novel methodological protocol for mapping the water status, i.e. the stem water potential (SWP), of citrus orchards was developed. Specifically, observed stem water potential (SWPₒbₛ) values and unmanned aerial vehicle multispectral information (i.e., vegetation indices, VIs, and spectral bands, SBs) were integrated to implement a twofold approach based on: (i) the spatial interpolation (SWPᵢₙₜ) of the SWPₒbₛ, and (ii) the stepwise regression models (SWPₚᵣₒₓy) between the SWPₒbₛ and the VIs (scenario 1) or between the SWPₒbₛ and the SBs (scenario 2). Then, the derived crop water status maps (SWPᵢₙₜ and SWPₚᵣₒₓy) were customized by applying an absolute (scientific-driven), a relative (quantile-driven), and an automated clustering (K-means) classification method. The accuracy of the proposed approach, evaluated by comparing SWPᵢₙₜ and SWPₚᵣₒₓy with SWPₒbₛ using linear regression models, showed reliable results, with average mean absolute error and root mean square error values ranging from 0.13 to 0.19 MPa and from 0.19 to 0.24 MPa, respectively. These results provide practical insights for identifying the spatial-temporal variability of the SWP of the citrus orchard under study. Additionally, the study highlights the importance of using a scientific-driven classification to support the adoption of precision irrigation criteria and decision-making process by non-expert users, as indicated by the assessment of the Silhouette index.

Why it matches plant phenotyping methodsUAVマルチスペクトル情報と空間補間・回帰モデルを統合し、柑橘樹の茎水ポテンシャルという生理状態を地図化する方法を開発・精度評価しており、植物フェノタイピング手法が中心である。

abstracta novel methodological protocol for mapping the water status, i.e. the stem water potential (SWP), of citrus orchards was developed.

Diagnosis of Citrus Greening Using Artificial Intelligence: A Faster Region-Based Convolutional Neural Network Approach with Convolution Block Attention Module-Integrated VGGNet and ResNet Models →

The vector-transmitted Citrus Greening (CG) disease, also called Huanglongbing, is one of the most destructive diseases of citrus. Since no measures for directly controlling this disease are available at present, current disease management integrates several measures, such as vector control, the use of disease-free trees, the removal of diseased trees, etc. The most essential issue in integrated management is how CG-infected trees can be detected efficiently. For CG detection, digital image analyses using deep learning algorithms have attracted much interest from both researchers and growers. Models using transfer learning with the Faster R-CNN architecture were constructed and compared with two pre-trained Convolutional Neural Network (CNN) models, VGGNet and ResNet. Their efficiency was examined by integrating their feature extraction capabilities into the Convolution Block Attention Module (CBAM) to create VGGNet+CBAM and ResNet+CBAM variants. ResNet models performed best. Moreover, the integration of CBAM notably improved CG disease detection precision and the overall performance of the models. Efficient models with transfer learning using Faster R-CNN were loaded on web applications to facilitate access for real-time diagnosis by farmers via the deployment of in-field images. The practical ability of the applications to detect CG disease is discussed.

Why it matches plant phenotyping methods柑橘の感染状態を画像から推定する深層学習手法を構築・比較し、検出性能を評価しているため、植物病害フェノタイピング手法が中心である。

abstractModels using transfer learning with the Faster R-CNN architecture were constructed and compared with two pre-trained Convolutional Neural Network (CNN) models, VGGNet and ResNet.

Dynamic Slicing and Reconstruction Algorithm for Precise Canopy Volume Estimation in 3D Citrus Tree Point Clouds →

Crop phenotyping data collection is the basis for precision agriculture and smart decision-making applications. Accurately obtaining the canopy volume of citrus trees is crucial for yield prediction, precise fertilization and cultivation management. To this end, we developed a dynamic slicing and reconstruction (DR) algorithm based on 3D point clouds. The algorithm dynamically slices nearby slices based on their proportional area change and density difference; for each slice point cloud, the average distance of each point from others is taken as the initial α value for the AS algorithm. This value is iteratively summed until it reconstructs the complete shape, allowing the volume of each slice shape to be determined. Compared with six point cloud-based reconstruction algorithms, the DR approach achieved the best results in removing perforations and lacunae (0.84) and exhibited volumetric consistency (1.53) that closely aligned with the growth pattern of citrus trees. The DR algorithm effectively addresses the challenges of adapting the thickness and number of canopy point cloud slices to the shape and size of the canopy in the ASBS and CHBS algorithms, as well as overcoming inaccuracies and incompleteness in reconstructed canopy models caused by limitations in capturing detailed features using the PCH algorithm. It offers improved adaptive ability, finer volume computations, better noise reduction, and anomaly removal.

Why it matches plant phenotyping methods柑橘樹の3D点群から樹冠体積という植物形態形質を推定するアルゴリズムを開発し、既存6手法と比較検証しており、表現型取得・抽出が研究の中心である。

abstractAccurately obtaining the canopy volume of citrus trees is crucial for yield prediction, precise fertilization and cultivation management. To this end, we developed a dynamic slicing and reconstruction (DR) algorithm based on 3D point clouds.

Performance Evaluation of YOLO Models in Plant Disease Detection →

Plant diseases significantly impact global agriculture, leading to substantial production losses and economic consequences. Timely disease detection can enhance crop yield, optimize resource utilization, reduce costs, and mitigate environmental effects, ultimately ensuring high-quality food production. Deep learning, specifically computer vision-based techniques, have proven invaluable in tasks like image classification, segmentation, and object detection. Deep Learning techniques such as You Only Look Once (YOLO) models are state of the art neural network algorithms used for accurate object detection. In this study, YOLOv5, YOLOv7 and YOLOv8 models were trained on CCL’20 dataset for citrus disease detection. Data augmentation techniques such as image translation, image scaling, flip, mosaic augmentations were implemented to improve the models’ performance during training phase. The model performance was evaluated using metric such as Mean Average Precision at 50% to 95% Intersection over Union score i.e. mAP@50-95. The results show that YOLOv8 model performs better than other variants and offers significant improvements over the benchmark performance from previous studies. The final hyper-parameter tuned model achieved 96.1% mAP@50-95 on testing data for citrus disease detection and mAP@50-95 of 95.3%, 96.0% and 97.0% for detection of Anthracnose, Melanose and Bacterial Brown Spot diseases, respectively. The trained model was able to detect single and multiple instances of same or different disease in an image showing the potential of recent YOLO models. The trained YOLOv8 model is deployed on Roboflow platform.

Why it matches plant phenotyping methods柑橘の画像から病害状態を推定するYOLOモデルを比較・評価し、データセット、性能指標、ハイパーパラメータ調整、展開まで扱っており、植物病害フェノタイピング手法が中心です。

abstractIn this study, YOLOv5, YOLOv7 and YOLOv8 models were trained on CCL’20 dataset for citrus disease detection.

Detection of Different Lemon Pre-Harvest Additive Treatments by Means an Electronic Nose Prototype →

This study was carried out with a low-cost electronic nose prototype based on eight metal oxide sensors (MQ) in order to characterize samples of lemons treated with 0.5% and 0.1% of sodium benzoate. The MQ sensors designed are sensitive to one or more chemicals to detect the presence of a variety of chemicals in the air. The sensor MQ135 detects ammonia, hydrogen sulphide and benzene. Signal data were studied to obtain a pattern recognition of rotten in lemon fruits. Network analysis was used to obtain a calibration of measures among the stage of lemons. In this article, an electronic nose prototype based on 8 MQ metal oxide sensors has been used in order to analyze and characterize different lemon varieties to which different chemical treatments have been applied in pre-harvest. PCA-based data analyzes were used to observe clusters in the data. Through the combined use of the data obtained by the nose and these Sequential Neural Networks (SNNs) a classification tool for lemon varieties and applied treatments has been obtained. It is shown the ability of this device to be used as a reliable discrimination method, in addition to providing low cost and optimization of time and expert resources.

Why it matches plant phenotyping methodsレモン果実の腐敗状態や品種・処理差を識別する電子鼻センサーと解析手法が研究の中心であり、植物器官の状態を測定するフェノタイピング手法に該当する。

abstractThis study was carried out with a low-cost electronic nose prototype based on eight metal oxide sensors (MQ) in order to characterize samples of lemons treated with 0.5% and 0.1% of sodium benzoate.

Fusion of fruit image processing and deep learning: a study on identification of citrus ripeness based on R-LBP algorithm and YOLO-CIT model. →

Citrus fruits are extensively cultivated fruits with high nutritional value. The identification of distinct ripeness stages in citrus fruits plays a crucial role in guiding the planning of harvesting paths for citrus-picking robots and facilitating yield estimations in orchards. However, challenges arise in the identification of citrus fruit ripeness due to the similarity in color between green unripe citrus fruits and tree leaves, leading to an omission in identification. Additionally, the resemblance between partially ripe, orange-green interspersed fruits and fully ripe fruits poses a risk of misidentification, further complicating the identification of citrus fruit ripeness. This study proposed the YOLO-CIT (You Only Look Once-Citrus) model and integrated an innovative R-LBP (Roughness-Local Binary Pattern) method to accurately identify citrus fruits at distinct ripeness stages. The R-LBP algorithm, an extension of the LBP algorithm, enhances the texture features of citrus fruits at distinct ripeness stages by calculating the coefficient of variation in grayscale values of pixels within a certain range in different directions around the target pixel. The C3 model embedded by the CBAM (Convolutional Block Attention Module) replaced the original backbone network of the YOLOv5s model to form the backbone of the YOLO-CIT model. Instead of traditional convolution, Ghostconv is utilized by the neck network of the YOLO-CIT model. The fruit segment of citrus in the original citrus images processed by the R-LBP algorithm is combined with the background segment of the citrus images after grayscale processing to construct synthetic images, which are subsequently added to the training dataset. The experiment showed that the R-LBP algorithm is capable of amplifying the texture features among citrus fruits at distinct ripeness stages. The YOLO-CIT model combined with the R-LBP algorithm has a Precision of 88.13%, a Recall of 93.16%, an F1 score of 90.89, a mAP@0.5 of 85.88%, and 6.1ms of average detection speed for citrus fruit ripeness identification in complex environments. The model demonstrates the capability to accurately and swiftly identify citrus fruits at distinct ripeness stages in real-world environments, effectively guiding the determination of picking targets and path planning for harvesting robots.

Why it matches plant phenotyping methods柑橘果実の成熟段階という植物器官の状態を、画像処理と深層学習で識別する手法を開発・評価しており、フェノタイピング手法が中心である。

abstractThis study proposed the YOLO-CIT (You Only Look Once-Citrus) model and integrated an innovative R-LBP (Roughness-Local Binary Pattern) method to accurately identify citrus fruits at distinct ripeness stages.

Dynamic Slicing and Reconstruction Algorithm for Precise Canopy Volume Estimation in 3D Citrus Tree Point Clouds →

Crop phenotyping data collection is the basis for precision agriculture and smart decisionmaking applications. In this study, a dynamic slicing and reconstruction canopy volume (DR) algorithm is proposed to accurately estimate citrus canopy volume. The algorithm dynamically slices nearby slices based on their proportional area change and density difference, subsequently conducting AS reconstruction and volume calculation for each slice using an iterative mean point spacing as the α-value. Compared with six point cloud-based reconstruction algorithms, the DR approach achieved the best results in removing perforations and lacunae (0.84) and exhibited volumetric consistency (1.53) that closely aligned with the growth pattern of citrus trees. The DR algorithm effectively addresses the challenges of adapting the thickness and number of canopy point cloud slices to the shape and size of the canopy in the ASBS and CHBS algorithms, as well as overcoming inaccuracies and incompleteness in reconstructed canopy models caused by limitations in capturing detailed features using the PCH algorithm. It offers improved adaptive ability, finer volume computations, better noise reduction, and anomaly removal. In conclusion, we recommend selecting appropriate operating environments for each algorithm based on their principles, geometric properties, volumetric values, running time, and linear relationships with one another to guide orchard mechanization and intelligent operation.

Why it matches plant phenotyping methods柑橘樹の3D点群から樹冠体積を推定するアルゴリズムを開発し、複数手法との比較検証を行っており、植物形質の取得・抽出が中心である。

abstracta dynamic slicing and reconstruction canopy volume (DR) algorithm is proposed to accurately estimate citrus canopy volume

Hybrid attention network for citrus disease identification →

Accurate identification and timely prevention of citrus diseases will effectively protect the interests of the citrus industry. However, the citrus disease identification models currently used in the industry have unsatisfactory performance due to low robustness. In this study, we comprehensively study the problem of citrus disease identification from both data and algorithm perspectives. In order to address the root cause of the negative impact of the actual complex orchard environment on the identification model in practical applications, an orchard context-based citrus disease dataset including Citrus yellow vein clearing virus (CYVCV), Canker, Brown spot, Melanose, Sooty mold, and healthy control is created. Data are collected from citrus leaf parts to maintain data uniformity. The model trained in this dataset can better adapt to the complex environment of the orchard, and thus more effectively serve the application in the production area. The key information for citrus disease identification is spot characteristic information, but due to the small size of the spots, it is difficult to focus and extract the characteristic information. In order to solve this problem, we studied that the representation of the features in the frequency domain dimension after the wavelet transform process is sparse, which is beneficial to improving the performance of the attention module, and proposed the frequency-domain attention network (FdaNet) to adaptively learn through the importance of feature information between different frequency domains changes the weight of each frequency domain during network inference. The effectiveness of FdaNet was demonstrated in experiments on citrus disease identification embedded in a ResNet backbone network. Next, according to the complex and diverse background of citrus disease data, a hybrid attention network (HaNet) is proposed to focus on multi-dimensional feature information. In HaNet, the frequency domain attention module is embedded into the channel attention network to enhance the channel scalar computing capability. In addition, in order to maximize the feature range extracted by the attention module in two dimensions, large convolution kernels are introduced in the backbone network to improve the effective perceptual field of the network. Moreover, we conducted research experiments using large convolution kernels of different sizes to further select sensory fields suitable for citrus disease feature extraction. Experimental results on the citrus disease dataset show that our proposed model achieves recognition accuracy of 98.83 % and 98.77 % on 50-layer and 101-layer networks respectively, both of which are better than other state-of-the-art models.

Why it matches plant phenotyping methods柑橘葉画像から病害状態を識別するデータセットと深層学習モデルを開発・評価しており、植物病害表現型の取得・推定手法が中心である。

abstractproposed the frequency-domain attention network (FdaNet) to adaptively learn through the importance of feature information between different frequency domains

Revolutionizing Plant Disease Detection: A Review of Deep Learning and Machine Learning Algorithms →

The food industry has led the agricultural economy of the state all India to prosperity. India has historically been the largest producing nation having identity of Agricultural Land. Grains , fruits , Vegetables , such as potatoes, oranges, Tomato ,sugarcane and other specially grains and cottons are the chief crops of the India. Citrus and cotton industries have been a driving force behind Maharashtra's impressive economic growth.. The situation has created job opportunities for many people, boosting the state's economic potential. To maintain the prosperity of citrus and cotton industries, Government has been concerned about disease control, labour cost, and global market. During the recent past, citrus canker and citrus greening, Black spot-n cotton has become serious threats to citrus in Maharashtra. Infection by these diseases weakens trees, leading to decline, mortality, lower yields, and decreased commercial value. Likewise, the farmers are concerned about costs from tree loss, scouting, and chemicals used in an attempt to control the disease. An automated detection system may help in prevention and, thus reduce the serious loss to the industries, farmers and Economy of country. This research aims to the development of disease detection with pattern recognition approaches for these diseases in crop. The detection approach consists of three major sub-systems, namely, image acquisition, image processing and pattern recognition. The imaging processing sub-system includes image preprocessing for background noise removal, leaf boundary detection and image feature extraction. Pattern recognition approaches will be use to classify samples among several different conditions on crops. In order to evaluate the classification approaches, results will be compared between classification methods for the different induvial fruits, vegetable, grains disease detection. Obtained results will help in demonstration of classification accuracy which is targeted as better than existing for proposed model as high as 97.00%. This study aimed to assess the potential of identifying plant diseases by examining visible signs on fruits and leaves. These data collection and initial knowledge acquisition is plan in offline approaches. By implementing this simple model, we can achieve a more favourable cost-to-production ratio compared to complex solutions.

Why it matches plant phenotyping methods植物の葉・果実の可視症状から病害状態を画像取得・画像処理・パターン認識で推定する手法の開発が中心であり、植物病害フェノタイピングに該当する。

abstractThis research aims to the development of disease detection with pattern recognition approaches for these diseases in crop.

Identifikasi Penyakit Daun Jeruk Siam Menggunakan Convolutional Neural Network (CNN) dengan Arsitektur EfficientNet →

Jeruk siam menjadi salah satu komoditas hortikultura yang memegang peranan utama dalam sektor pertanian Indonesia dengan jumlah produksi yang mencapai 2 juta ton setiap tahunnya. Namun, produksi jeruk siam rentan terhadap serangan hama dan penyakit, terutama pada bagian daun. Penyakit yang umum terjadi termasuk Blackspot Leaf, Canker Leaf, Greening Leaf, Powdery Mildew, dan Citrus Leafminer. Pada umunya identifikasi penyakit pada tanaman jeruk dilakukan secara manual sehingga penentuan penyakit cenderung subyektif. Oleh karena itu, diperlukan solusi otomatis dalam mendeteksi penyakit pada daun jeruk. Tujuan penelitian yaitu untuk mengidentifikasi penyakit yang menyerang daun jeruk menggunakan metode deep learning yaitu CNN dengan arsitektur EfficientNetB3. Dataset yang digunakan adalah citra penyakit daun jeruk yang diambil langsung dari kebun jeruk yang dibagi menjadi 6 kelas seperti pada penyakit yang disebutkan di atas. Hasil penelitian menggunakan skenario epoch 10 dengan optimizer Adam memperoleh hasil akurasi terbaik yaitu 0,98 (98%).

Why it matches plant phenotyping methods葉の画像から柑橘の病害状態をCNNで自動識別する手法が研究の中心であり、植物病害フェノタイピングに該当する。

titleIdentifikasi Penyakit Daun Jeruk Siam Menggunakan Convolutional Neural Network (CNN) dengan Arsitektur EfficientNet

Design of citrus peel defect and fruit morphology detection method based on machine vision →

Identifying defects in citrus peels and analyzing fruit morphology are two core challenges in citrus quality inspection. In order to more accurately identify minor defects on citrus peels, we proposed a detection model Yolo-FD (Yolo for defects). The model was based on the Yolov5 network framework, and the backbone network embedded the Three-dimensional Coordinate Attention (TDCA) mechanism innovatively designed in this study. It accurately captured the subtle changes and feature associations of the target in spatial location, significantly enhancing the model's ability to perceive defects in fruit peels. Moreover, we employed a simplified Bidirectional Weighted Feature Pyramid Network (BiFPN) in the model to achieve cross-scale connections and improve the feature fusion ability of the model. At the same time, Contextual Transformer block (COT) was introduced into Neck network and the CoT3 module was built to fully capture the static and dynamic contextual information in the citrus defects images and enhance the expression of the feature map. Through this series of improvement methods, missed detections and false detections caused by small targets were effectively reduced. Fruit morphology detection was combined with the Partice Swarm Optimized Extreme Learning Machine (PSO-ELM) model to determine whether the citrus fruit morphology was well-formed, using the symmetry index, roundness and tilt angle of the citrus as input parameters. The experimental results indicated that the mean average precision of the Yolo-FD model is 98.7 % (mAP-0.5). Compared with Yolov5s, Yolov7-tiny, and Yolov8n, the mAP was improved by 1.4 %, 1.5 %, and 0.5 % respectively. Its average detection time for a single frame image on the server was 19.5 ms. And the PSO-ELM model achieved a fruit morphology detection accuracy of 91.42 %, a coefficient of determination of 0.9044, and a mean squared error of 0.8497. The research results met the accuracy and real-time requirements for citrus sorting on the production line, and could provide an effective solution for citrus grading and quality assessment.

Why it matches plant phenotyping methods柑橘果实形态（对称性、圆度、倾角）を画像から推定する手法を開発・評価しており、果実形態の取得・抽出が中心的な技術貢献である。

abstractFruit morphology detection was combined with the Partice Swarm Optimized Extreme Learning Machine (PSO-ELM) model to determine whether the citrus fruit morphology was well-formed, using the symmetry index, roundness and tilt angle of the citrus as input parameters.

A high‐throughput analysis of high‐resolution X‐ray CT images of stems of olive and citrus plants resistant and susceptible to Xylella fastidiosa →

The bacterial plant pathogen Xylella fastidiosa causes disease in several globally important crops. However, some cultivars harbour reduced bacterial loads and express few symptoms. Evidence considering plant species in isolation suggests xylem structure influences cultivar susceptibility to X. fastidiosa. We test this theory more broadly by analysing high‐resolution synchrotron X‐ray computed tomography of healthy and infected plant vasculature from two taxonomic groups containing susceptible and resistant varieties: two citrus cultivars (sweet orange cv. Pera, tangor cv. Murcott) and two olive cultivars (Koroneiki, Leccino). Results found the susceptible plants had more vessels than resistant ones, which could promote within‐host pathogen spread. However, features associated with resistance were not shared by citrus and olive. While xylem vessels in resistant citrus stems had comparable diameters to those in susceptible plants, resistant olives had narrower vessels that could limit biofilm spread. And while differences among olive cultivars were not detected, results suggest greater vascular connectivity in resistant compared to susceptible citrus plants. We hypothesize that this provides alternate flow paths for sustaining hydraulic functionality under infection. In summary, this work elucidates different physiological resistance mechanisms between two taxonomic groups, while supporting the existence of an intertaxonomical metric that could speed up the identification of candidate‐resistant plants.

Why it matches plant phenotyping methods高解像度X線CT画像から植物の木部血管形態・連結性を抽出し、耐病性候補指標として比較することが研究の中心であり、植物状態の画像ベース表現型解析に該当する。

titleA high‐throughput analysis of high‐resolution X‐ray CT images of stems of olive and citrus plants resistant and susceptible to Xylella fastidiosa

Plant Disease Detection Using YOLOv7 Algorithm →

Abstract: Plant diseases pose a serious danger to global food security and can result in huge financial losses for the agricultural sector. Earlier Plant disease detection and precise diagnosis are essential for putting management measures into place. Recent advancements in computer vision techniques have demonstrated encouraging outcomes in automating activities related to illness identification. This study uses the You Only Look Once (YOLOv7) object identification algorithm to present a novel method for plant disease diagnosis. The primary goal of this research is to create a reliable and effective system that can quickly and reliably identify plant diseases. YOLOv7, a highly accurate and speedy algorithm, will serve as the main underpinning for detection. The project's primary goal is to train the Yolov7 model to identify distinct citrus plant illnesses by using a large dataset that includes pictures of both healthy and diseased plants. This project categorizes leaf images recorded from a file or webcam into four categories: healthy, greening, blackspot, and canker. Early disease prediction allows farmers to take required security measures for their plants.

Why it matches plant phenotyping methodsYOLOv7による葉画像からの植物病害状態の分類手法が研究の中心であり、植物の病徴・健全性を直接推定している。

abstractThis study uses the You Only Look Once (YOLOv7) object identification algorithm to present a novel method for plant disease diagnosis.

Analysis of the peel structure of different Citrus spp. via light microscopy, SEM and μCT with manual and automatic segmentation. →

The peels of lime, lemon, pomelo and citron are investigated at macroscopic and microscopic level. The structural composition of the peels is compared and properties such as peel thickness, proportion of flavedo, density and proportion of intercellular spaces are determined. μCT images are used to visualize vascular bundles and oil glands. SEM images provide information about the appearance of the cellular tissue in the outer flavedo and inner albedo. The proportion of intercellular spaces is quantitatively determined by manual and software-assisted analysis (ilastik). While there are macroscopic differences in the fruits, they differ only slightly in the orientation of the vascular bundles and the arrangement of the oil glands. However, in peel thickness and flavedo thickness, the fruit peels differ significantly from each other. There are no significant differences between the two analysis methods used, although the use of ilastik is preferred due to time reduction of up to 70%. The large amount of intercellular spaces in the albedo but also the denser flavedo both have a mechanical protective function to prevent damage to the fruit. In addition, the entire peel structure is mechanically reinforced by vascular bundles. This combination of penetration protection (flavedo) and energy dissipation (albedo) makes Citrus spp. peels a promising inspiration for technical material systems.

Why it matches plant phenotyping methods柑橘果皮の構造形質を光学・SEM・μCT画像から取得し、手動法とソフトウェア支援セグメンテーションを比較・検証しているため、植物形質取得法が中心である。

abstractproperties such as peel thickness, proportion of flavedo, density and proportion of intercellular spaces are determined.

A small neural network deployed on edge devices for detecting citrus Huanglongbing. →

Abstract Citrus Huanglongbing (HLB) poses a significant threat to the profitability of the citrus industry worldwide. In traditional agricultural practices, manually identifying citrus trees infected with HLB based on certain leaf characteristics is time-consuming, subjective, and inefficient. The initial automatic identification of citrus Huanglongbing (HLB) relies on traditional image processing and machine learning algorithms, exhibiting low accuracy and slow processing speed. In order to enhance both the detection accuracy and speed, researchers have introduced deep learning methods based on neural networks for the identification of citrus HLB. However, the neural network models currently used for citrus leaf HLB identification have large parameter sizes, high deployment costs, and require high computational power, making them unsuitable for deployment on edge devices for field detection. Therefore, in order to promptly detect and address diseased plants, improve farmers' agricultural operational efficiency, ensure the accessibility of deep learning in small-scale agriculture, and address the need for cost-effective measures, there is an urgent need for a low-cost deep learning framework. Therefore, we compared the performance of several commonly used deep convolutional neural networks in industry for citrus Huanglongbing (HLB) detection. We constructed image classification networks based on AlexNet, ResNet, MobileNet-V1, and MobileNet-V3, and evaluated the network models based on model size, parameter count, and classification performance. As a result, we proposed a deep learning-based method for detecting citrus HLB. This method has a small model parameter count, low computational cost, fast detection speed, and high detection accuracy. It can be deployed on edge devices or other embedded devices. This method has a small model parameter count, fast detection speed, and high accuracy. The classification task is achieved by training the overall feature extraction network and the classification network at the network's tail on the constructed training set. The actual detection results show that the detection accuracy for healthy citrus leaves reaches 99.02%, and for HLB-infected leaves, the detection accuracy reaches 99.07%. The overall accuracy is 99.04%. Both recall and precision rates are excellent, meeting the precision requirements for on-site detection.

Why it matches plant phenotyping methods柑橘葉の画像からHLB感染状態を推定する深層学習手法を構築・比較し、精度や計算性能を評価しているため、植物病害表現型の取得・推定が中心である。

abstractWe constructed image classification networks based on AlexNet, ResNet, MobileNet-V1, and MobileNet-V3, and evaluated the network models based on model size, parameter count, and classification performance.

Automated classification of citrus disease on fruits and leaves using convolutional neural network generated features from hyperspectral images and machine learning classifiers →

Citrus black spot (CBS) is a fungal disease caused by Phyllosticta citricarpa that poses a quarantine threat and can restrict market access to fruits. It manifests as lesions on the fruit surface and can result in premature fruit drops, leading to reduced yield. Another significant disease affecting citrus is canker, which is caused by the bacterium Xanthomonas citri subsp. citri (syn. X. axonopodis pv. citri); it causes economic losses for growers due to fruit drops and blemishes. Early detection and management of groves infected with CBS or canker through fruit and leaf inspection can greatly benefit the Florida citrus industry. However, manual inspection and classification of disease symptoms on fruits or leaves are labor-intensive and time-consuming processes. Therefore, there is a need to develop a computer vision system capable of autonomously classifying fruits and leaves, expediting disease management in the groves. This paper aims to demonstrate the effectiveness of convolutional neural network (CNN) generated features and machine learning (ML) classifiers for detecting CBS infected fruits and leaves with canker symptoms. A custom shallow CNN with radial basis function support vector machine (RBF SVM) achieved an overall accuracy of 92.1% for classifying fruits with CBS and four other conditions (greasy spot, melanose, wind scar, and marketable), and a custom Visual Geometry Group 16 (VGG16) with the RBF SVM classified leaves with canker and four other conditions (control, greasy spot, melanoses, and scab) at an overall accuracy of 93%. These preliminary findings demonstrate the potential of utilizing hyperspectral imaging (HSI) systems for automated classification of citrus fruit and leaf diseases using shallow and deep CNN-generated features, along with ML classifiers.

Why it matches plant phenotyping methods柑橘の果実・葉に現れる病徴をハイパースペクトル画像とCNN/機械学習で自動分類する方法が研究の中心であり、植物病害状態の表現型推定に該当する。

abstractTherefore, there is a need to develop a computer vision system capable of autonomously classifying fruits and leaves, expediting disease management in the groves.

Comparison of leaf chlorophyll content retrieval performance of citrus using FOD and CWT methods with field-based full-spectrum hyperspectral reflectance data →

Citrus is one of the most economically valuable fruit trees in the world, for which leaf chlorophyll content (LCC) serves as a crucial indicator for evaluating its growth and health status. However, the quantitative estimation of LCC using remote-sensing techniques is still challenging owing to unclear sensitive spectral ranges, baseline drift and overlapping spectrum peaks. To resolve these issues, we clarified the spectral response characteristics of citrus LCC using fractional-order derivatives (FOD) and continuous wavelet transform (CWT) methods to determine its sensitive spectral range with in situ full-spectrum leaf hyperspectral data. We proposed a novel method for estimating the LCC of citrus by combining an ensemble learning regression model based on Hyperopt optimization (H-ELR) with partial least squares regression (PLSR) using the ultra-high dimensional feature variables produced by dual- and tri-band combination strategies. We evaluated the retrieval performance of LCC between the FOD- and CWT-based optimal spectral feature variables. Besides, we further examined the feasibility of improving the estimation accuracy of LCC by the combination of their optimal feature variables. Finally, we evaluated the effect of the spectral curve trend changes and different dimensional spectral features on LCC estimation. The results showed that: (1) The FOD- and CWT-based methods improved the correlation between original spectral reflectance and LCC, with the correlation coefficient increasing by 0.046 and 0.054, respectively. We confirmed that 425–740 nm is the optimal spectral range for LCC estimation. (2) We found that the 0.9 order derivative Tri-band index (TBI2 (R₅₇₁, R₁₆₉₇, R₇₄₀)) constructed by combining the sensitive spectral bands of leaf water content achieved a high-precision LCC estimation (R² = 0.876). In addition, the MERIS terrestrial chlorophyll index (MTCI) constructed based on the scale6 of CWT also gained good retrieval accuracy (R² = 0.806). (3) We demonstrated that a combination of FOD-based and CWT-based sensitive spectral features improves estimation accuracies (R² = 0.891) of LCC and revealed that the reflectance peaks and slope peaks at 550 and 750 nm are essential variables for predicting the citrus LCC. (4) The combination of PLSR and H-ELR model provided a good retrieval performance of citrus LCC with the kurtosis (γ = 3.2) and skewness (Sk = 0.066) of the residual prediction values. Our proposed method can provide a scientific basis for estimating LCC and other physiological parameters of citrus and other crop types, which is also important to optimize agricultural management practices and improve crop yield.

Why it matches plant phenotyping methods柑橘葉のクロロフィル含量という植物生理形質を、ハイパースペクトル反射データとFOD/CWT・回帰モデルで推定する手法の開発および性能比較が中心である。

abstractWe proposed a novel method for estimating the LCC of citrus by combining an ensemble learning regression model based on Hyperopt optimization (H-ELR) with partial least squares regression (PLSR)

A colorimetric sensor array combined with a hybrid feature selection approach for discrimination of citrus infested by Bactrocera dorsalis (Hendel) →

Changes in citrus volatile organic compounds (VOCs) induced by Bactrocera dorsalis (Hendel) infestation can serve as characteristic identifiers for non-destructive detection of infested citrus. This study proposed an innovative method combining colorimetric sensor array (CSA) technology with machine learning algorithms for the discrimination of B. dorsalis infestation in citrus. Gas chromatography-mass spectrometry (GC-MS) analysis identified key VOCs, including d-limonene, linalool, and decanal, as infestation markers. Subsequently, various porphyrin and metalloporphyrin dyes exhibiting sensitivity to these VOCs were selected to construct the CSA. To enhance detection accuracy, a hybrid feature selection method integrating ReliefF and Particle Swarm Optimization (PSO) was implemented. Subsequently, the optimized features subsets were utilized to develop classification models. Specifically, a binary classification model employing the K Nearest Neighbor (KNN) algorithm achieved a high accuracy of 93.89 % in distinguishing between healthy and infected citrus. Furthermore, a multi-class classification model using KNN was developed to differentiate among invasive, incubation, and infestation stages, attaining a remarkable accuracy of 97.78 %. This approach presents a promising solution for early detection of B. dorsalis infestation in citrus.

Why it matches plant phenotyping methods柑橘の害虫感染状態を、カラーセンサーアレイと機械学習で非破壊的に識別する測定・解析法が研究の中心であり、植物の病害虫状態の表現型推定に該当する。

abstractThis study proposed an innovative method combining colorimetric sensor array (CSA) technology with machine learning algorithms for the discrimination of B. dorsalis infestation in citrus.

Characterisation and optical detection of puffy Satsuma mandarin →

Puffiness is one of the dominant postharvest disorders in easy-peeling citrus cultivars. In this study, the structural changes between healthy and puffy Satsuma mandarin were investigated and the potential of using optical methods for disorder detection was explored. To gain more insight in this disorder, the external appearance and internal quality attributes were first compared between healthy and puffy Iwasaki Satsuma mandarins at three harvest times. Although no consistent differences were observed in the appearance of fruits, the soluble solids content and Brix minus acid values in puffy mandarin were found to be higher compared to the corresponding healthy fruit. The structural properties of the flavedo and albedo tissue layer in the peel were quantified from X-ray CT scans. Whilst no differences were observed in the size of the oil glands in the flavedo, the pore size in the albedo of puffy mandarin was found to be larger with later harvest. The bulk optical properties of the intact fruit were estimated from laser scatter images with a metamodel calibrated on optical phantoms. The reduced scattering coefficient (μₛ') for the intact fruit was found to be lower in puffy mandarin relative to healthy fruit. The distinction between healthy and puffy mandarin based on μₛ' was further validated on Goku Wase Satsuma mandarin. The results obtained indicate that healthy and puffy mandarin can be separated well based on their μₛ' at all the selected wavelengths. This provides a basis for the non-destructive optical detection of puffing disorder at an early stage.

Why it matches plant phenotyping methods柑橘果実の生理・構造状態（puffiness）を光学計測とX線CTで非破壊検出する方法を開発・検証しており、植物フェノタイピング手法が中心である。

abstractthe potential of using optical methods for disorder detection was explored

A Metabolomics Assay to Diagnose Citrus Huanglongbing Disease and to Aid in Assessment of Treatments to Prevent or Cure Infection. →

Citrus greening disease, or Huanglongbing (HLB), has devastated citrus crops globally in recent years. The causal bacterium, ' Candidatus Liberibacter asiaticus', presents a sampling issue for qPCR diagnostics and results in a high false negative rate. In this work, we compared six metabolomics assays to identify HLB-infected citrus trees from leaf tissue extracted from 30 control and 30 HLB-infected trees. A liquid chromatography-mass spectrometry-based assay was most accurate. A final partial least squares-discriminant analysis (PLS-DA) model was trained and validated on 690 leaf samples with corresponding qPCR measures from three citrus varieties (Rio Red grapefruit, Hamlin sweet orange, and Valencia sweet orange) from orchards in Florida and Texas. Trees were naturally infected with HLB transmitted by the insect vector Diaphorina citri . In a randomized validation set, the assay was 99.9% accurate to classify diseased from nondiseased samples. This model was applied to samples from trees receiving plant defense-inducer compounds or biological treatments to prevent or cure HLB infection. From two trials, HLB-related metabolite abundances and PLS-DA scores were tracked longitudinally and compared with those of control trees. We demonstrate how our assay can assess tree health and the efficacy of HLB treatments and conclude that no trialed treatment was efficacious.

Why it matches plant phenotyping methods柑橘葉の代謝プロファイルからHLB感染状態を推定する診断法を比較・開発し、多数サンプルで検証している。感染植物の健康状態・病害状態を測定する方法が中心であり、単なる分子診断や routine assay ではない。

abstractIn this work, we compared six metabolomics assays to identify HLB-infected citrus trees from leaf tissue extracted from 30 control and 30 HLB-infected trees.

Detection of mealybug infestation on the Khasi Mandarin orange plant using electronic nose →

Mealybugs pose a serious threat to fruit crops leading to premature leaf and fruit drops which severely affects the yield as well as the quality. Primarily pest detection is done with the help of human/animal scouting, which is cumbersome and prone to error. This paper studies the feasibility of using an electronic nose (E-Nose) for detecting mealybug infestation in Khasi Mandarin orange plants. Plants normally release volatile organic compounds (VOCs) which can act as biomarkers for specific stresses affecting the plant. These VOCs can be analyzed to diagnose the plant. VOCs emanating from leaf samples of both infested and healthy plants were analyzed using a custom-made E-Nose system containing an array of commercially available gas sensors. Dimensionality reduction techniques using principle component analysis, and linear discriminant analysis and optimized classification algorithms like support vector machine and random forest were employed to check for the discriminating capability of the E-Nose system. The technique successfully classified samples belonging to infested and healthy categories in both the classifiers with accuracies of 95.66% and 96.70%.

Why it matches plant phenotyping methods植物葉由来のVOCをE-Noseで測定し、健全・食害状態を分類する方法が研究の中心であり、植物のストレス状態を推定する実質的なセンシング手法である。

abstractThis paper studies the feasibility of using an electronic nose (E-Nose) for detecting mealybug infestation in Khasi Mandarin orange plants.

Design and test of Kinect-based variable spraying control system for orchards. →

Target detection technology and variable-rate spraying technology are key technologies for achieving precise and efficient pesticide application. To address the issues of low efficiency and high working environment requirements in detecting tree information during variable spraying in orchards, this study has designed a variable spraying control system. The system employed a Kinect sensor to real-time detect the canopy volume of citrus trees and adjusted the duty cycle of solenoid valves by pulse width modulation to control the pesticide application. A canopy volume calculation method was proposed, and precision tests for volume detection were conducted, with a maximum relative error of 10.54% compared to manual measurements. A nozzle flow model was designed to determine the spray decision coefficient. When the duty cycle ranged from 30% to 90%, the correlation coefficient of the flow model exceeded 0.95, and the actual flow rate of the system was similar to the theoretical flow rate. Field experiments were conducted to evaluate the spraying effectiveness of the variable spraying control system based on the Kinect sensor. The experimental results indicated that the variable spraying control system demonstrated good consistency between the theoretical spray volume and the actual spray volume. In deposition tests, compared to constant-rate spraying, the droplets under the variable-rate mode based on canopy volume exhibited higher deposition density. Although the amount of droplet deposit and coverage slightly decreased, they still met the requirements for spraying operation quality. Additionally, the variable-rate spray mode achieved the goal of reducing pesticide use, with a maximum pesticide saving rate of 57.14%. This study demonstrates the feasibility of the Kinect sensor in guiding spraying operations and provides a reference for their application in plant protection operations.

Why it matches plant phenotyping methodsKinectセンサーによる柑橘樹冠体積の推定法を開発・精度検証し、可変散布制御へ応用しており、植物形態計測が中心的な技術貢献である。

abstractThe system employed a Kinect sensor to real-time detect the canopy volume of citrus trees

Denoising Diffusion Probabilistic Models and Transfer Learning for citrus disease diagnosis. →

Problems Plant Disease diagnosis based on deep learning mechanisms has been extensively studied and applied. However, the complex and dynamic agricultural growth environment results in significant variations in the distribution of state samples, and the lack of sufficient real disease databases weakens the information carried by the samples, posing challenges for accurately training models. Aim This paper aims to test the feasibility and effectiveness of Denoising Diffusion Probabilistic Models (DDPM), Swin Transformer model, and Transfer Learning in diagnosing citrus diseases with a small sample. Methods Two training methods are proposed: The Method 1 employs the DDPM to generate synthetic images for data augmentation. The Swin Transformer model is then used for pre-training on the synthetic dataset produced by DDPM, followed by fine-tuning on the original citrus leaf images for disease classification through transfer learning. The Method 2 utilizes the pre-trained Swin Transformer model on the ImageNet dataset and fine-tunes it on the augmented dataset composed of the original and DDPM synthetic images. Results and conclusion The test results indicate that Method 1 achieved a validation accuracy of 96.3%, while Method 2 achieved a validation accuracy of 99.8%. Both methods effectively addressed the issue of model overfitting when dealing with a small dataset. Additionally, when compared with VGG16, EfficientNet, ShuffleNet, MobileNetV2, and DenseNet121 in citrus disease classification, the experimental results demonstrate the superiority of the proposed methods over existing approaches to a certain extent.

Why it matches plant phenotyping methods柑橘葉画像から病害状態を推定する画像解析手法を開発・比較し、データ拡張と分類性能を検証しているため、植物フェノタイピング手法が中心である。

abstractThis paper aims to test the feasibility and effectiveness of Denoising Diffusion Probabilistic Models (DDPM), Swin Transformer model, and Transfer Learning in diagnosing citrus diseases with a small sample.

CitrusUAT: A dataset of orange Citrus sinensis leaves for abnormality detection using image analysis techniques. →

Around the world, citrus production and quality are threatened by diseases caused by fungi, bacteria, and viruses. Citrus growers are currently demanding technological solutions to reduce the economic losses caused by citrus diseases. In this context, image analysis techniques have been widely used to detect citrus diseases, extracting discriminant features from an input image to distinguish between healthy and abnormal cases. The dataset presented in this article is helpful for training, validating, and comparing citrus abnormality detection algorithms. The data collection comprises 953 color images taken from the orange leaves of Citrus sinensis (L.) Osbeck species. There are 12 nutritional deficiencies and diseases supporting the development of automatic detection methods that can reduce economic losses in citrus production.

Why it matches plant phenotyping methods柑橘葉の異常・病害を画像で検出するためのデータセットで、検出アルゴリズムの訓練・検証・比較を主目的としており、植物フェノタイピング手法が中心です。

abstractThe dataset presented in this article is helpful for training, validating, and comparing citrus abnormality detection algorithms.

Parallel Fusion Neural Network Considering Local and Global Semantic Information for Citrus Tree Canopy Segmentation →

Existing convolutional neural network (CNN)-based methods usually tend to ignore the contextual information for citrus tree canopy segmentation. Although popular Transformer models are helpful in extracting global semantic information, they ignore the edge details between citrus tree canopies and the background. To address these issues, we propose a parallel fusion neural network considering both local and global semantic information for citrus tree canopy segmentation from 3D data, which are derived by unmanned aerial vehicle (UAV) mapping. In the feature extraction stage, a parallel architecture, concatenated by EfficientNet-V2 and CSwin Transformer, is used to extract local and global information of citrus trees. In the feature fusion stage, we design a coordinate attention-based fusion module to retain the contextual information and local edge details of citrus tree canopies. Additionally, to exaggerate the exclusivity between tree canopies and complex backgrounds, 3D data incorporating RGB imagery and canopy height model derived by UAV photogrammetry are generated for citrus tree canopy segmentation. Experimental results indicate that the proposed method performs considerably better than methods based only on CNN or Transformer models, and is superior to state-of-the-art methods (e.g., the highest mIoU score of 93.46%).

Why it matches plant phenotyping methods柑橘樹冠を対象とする3D画像セグメンテーション手法の開発・比較が中心であり、樹冠という植物形態状態を抽出するため、植物フェノタイピング手法として適格。

abstractwe propose a parallel fusion neural network considering both local and global semantic information for citrus tree canopy segmentation from 3D data

The reliability of Unmanned Aerial Vehicles (UAVs) equipped with multispectral cameras for estimating chlorophyll content, plant height, canopy area, and fruit total number of Lemons (Citrus limon) →

Monitoring lemon production requires appropriate and efficient technology. The use of UAVs can addressed these challenges. The purpose of this study was to determine the best vegetation indices (VIs) for estimating chlorophyll content, plant height (PH), canopy area (CA), and fruit total numberas (FTN). CCM 200 was used as a tool to measure the chlorophyll content index (CCI), the number of fruits was measured by hand-counter, and other variables were recorded in meters. The UAV used was a Phantom 4 with a multispectral camera capable of capturing five different bands. The VIs was obtained via analysis of digital numbers generated by the multispectral camera. Then, the VIs was correlated with the CCI, PH, CA and FTN. VIs tested included the following: the normalized difference vegetation index (NDVI), the normalized difference vegetation index-green (NDVIg), the normalized different index (NDI), green minus red (GMR), simple ratio (SR), the Visible Atmospherically Resistant Index (VARI), normalized difference red edge (NDRE), simple ratio red-edge (SR RE ), the simple ratio vegetation index (SR VI ), and the Canopy Chlorophyll Content Index (CCCI). The best model for predicting CCI was obtained using the NDVIg (R 2 =0.8480; RMSE=6.1665 and RRMSE=0.0908). Meanwhile, SR turned out to be the best model for predicting PH (R 2 =0.8266; RMSE=15.6432 and RRMSE=0.0883), CA (R 2 =0.6886; RMSE= 0.8826 and RRMSE=0.1907), and FTN (R 2 =0.6850; RMSE=24.5574 and RRMSE=0.3503). The implication of these results for future activities includes establishing early monitoring and evaluation systems for lemon yield and production. This model was developed and tested in this specific location and under these environmental conditions.

Why it matches plant phenotyping methodsUAVマルチスペクトル画像と植生指数を用いてレモンのクロロフィル、草丈、樹冠面積、果実数を推定し、モデル性能を評価することが研究の中心である。

abstractThe purpose of this study was to determine the best vegetation indices (VIs) for estimating chlorophyll content, plant height (PH), canopy area (CA), and fruit total numberas (FTN).

Early diagnosis and mechanistic understanding of citrus Huanglongbing via sun-induced chlorophyll fluorescence →

Citrus Huanglongbing (HLB) is a serious disease that is devastating the citrus industry worldwide. The infection of HLB causes a significant impact on the photosynthesis and growth of citrus but the early diagnosis of HLB is challenging. This study aims to propose an effective method for early HLB diagnosis and mechanistic understanding via sun-induced chlorophyll fluorescence (SIF) during the whole growth period of citrus. SIF signals and net photosynthetic rate (A) of healthy, asymptomatic HLB (aHLB) and symptomatic HLB (sHLB) citrus leaf samples were acquired throughout the whole citrus growth period. Seasonal variations of the peak SIF yield at 687 nm (SIFY₆₈₇) and 741 nm (SIFY₇₄₁) and A were investigated. Leaf chlorophyll a content (Cₐ), chlorophyll b content (Cb), and carotenoids content (Cₓc) were also measured by spectrophotometry in the laboratory. It was found that the A of the aHLB was close to that of sHLB, which was significantly different from the healthy leaves at the physiological fruit dropping (PFD) period. The ratio of chlorophyll content (Chl) to Cₓc (Chl/Cₓc) was evidently different between aHLB and healthy leaves at the late PFD and early fruit expanding (FE) stage, which could be a potential biomarker to identify HLB in the incubation period. The classification result showed PFD was the optimal stage for HLB diagnosis and the index based on SIFY₇₄₁ was recommended (overall accuracy > 90 %). Moreover, HLB had a great influence on the relationship between A and SIF at the PFD stage, and the coefficients of determination (R²) of the univariate linear regression model between A and SIFY₆₈₇ and SIFY₇₄₁ decreased from 0.81 and 0.72 to 0.21 and 0.17, respectively. These results demonstrate the HLB effect on the relationship between A and SIF, and that the SIF could be a proxy to citrus photosynthesis and further used to diagnose early HLB in situ.

Why it matches plant phenotyping methods植物の光合成状態を測定するSIFを用いたHLB早期診断法の提案・評価が研究の中心であり、病害状態という植物表現型を直接推定している。

abstractThis study aims to propose an effective method for early HLB diagnosis and mechanistic understanding via sun-induced chlorophyll fluorescence (SIF) during the whole growth period of citrus.

Implementing deep‐learning techniques for accurate fruit disease identification →

To overcome the problems of manual identification of fruit disease, this work proposes a deep‐learning model to analyse fruit images to detect diseases in the fruit. We are proposing here a convolutional neural network (CNN)‐based model for fruit disease classification. By including many layers, the proposed CNN model extracts numerous features from the fruit, deals with the large data set and finally evaluates it. With the MobileNetv2 model, the disease prediction accuracy for papaya, guava and citrus was 99.4%, 98.8% and 95.8% and the recall values were 99.4%, 98.8% and 93.8%, respectively. With VGG16, the disease prediction accuracy for papaya, guava and citrus was 97.7%, 99.6% and 94.2% and the recall values were 96.5%, 99.6% and 89.2%, respectively. Finally, with DenseNet121, the disease prediction accuracy for papaya, guava and citrus was 99.4%, 97.6% and 99.2%, and the recall values were 98.8%, 97.6% and 99.2%, respectively.

Why it matches plant phenotyping methods果実画像から病害状態をCNNで推定する手法の提案・評価が中心であり、植物の病徴を直接観測する画像ベース表現型解析に該当する。

abstractthis work proposes a deep‐learning model to analyse fruit images to detect diseases in the fruit

Detection of citrus diseases in complex backgrounds based on image-text multimodal fusion and knowledge assistance. →

Diseases pose a significant threat to the citrus industry, and the accurate detection of these diseases represent key factors for their early diagnosis and precise control. Existing diagnostic methods primarily rely on image models trained on vast datasets and limited their applicability due to singular backgrounds. To devise a more accurate, robust, and versatile model for citrus disease classification, this study focused on data diversity, knowledge assistance, and modal fusion. Leaves from healthy plants and plants infected with 10 prevalent diseases (citrus greening, citrus canker, anthracnose, scab, greasy spot, melanose, sooty mold, nitrogen deficiency, magnesium deficiency, and iron deficiency) were used as materials. Initially, three datasets with white, natural, and mixed backgrounds were constructed to analyze their effects on the training accuracy, test generalization ability, and classification balance. This diversification of data significantly improved the model's adaptability to natural settings. Subsequently, by leveraging agricultural domain knowledge, a structured citrus disease features glossary was developed to enhance the efficiency of data preparation and the credibility of identification results. To address the underutilization of multimodal data in existing models, this study explored semantic embedding methods for disease images and structured descriptive texts. Convolutional networks with different depths (VGG16, ResNet50, MobileNetV2, and ShuffleNetV2) were used to extract the visual features of leaves. Concurrently, TextCNN and fastText were used to extract textual features and semantic relationships. By integrating the complementary nature of the image and text information, a joint learning model for citrus disease features was achieved. ShuffleNetV2 + TextCNN, the optimal multimodal model, achieved a classification accuracy of 98.33% on the mixed dataset, which represented improvements of 9.78% and 21.11% over the single-image and single-text models, respectively. This model also exhibited faster convergence, superior classification balance, and enhanced generalization capability, compared with the other methods. The image-text multimodal feature fusion network proposed in this study, which integrates text and image features with domain knowledge, can identify and classify citrus diseases in scenarios with limited samples and multiple background noise. The proposed model provides a more reliable decision-making basis for the precise application of biological and chemical control strategies for citrus production.

Why it matches plant phenotyping methods柑橘葉の病害状態を画像から分類するマルチモーダル手法の開発が研究の中心であり、植物の病害表現型を直接推定している。

abstractTo devise a more accurate, robust, and versatile model for citrus disease classification, this study focused on data diversity, knowledge assistance, and modal fusion.

Early detection of citrus anthracnose caused by Colletotrichum gloeosporioides using hyperspectral imaging →

Citrus fruit are susceptible to Colletotrichum gloeosporioides infestation during postharvest and shelf storage. Early and accurate detection of citrus anthracnose is conducive for carrying out targeted pesticide control and mitigating the potential spread of the disease. An early citrus anthracnose detection method using hyperspectral imaging and machine learning techniques is proposed. The hyperspectral data of sound citrus fruits were first collected and served as healthy samples, which were then inoculated with C. gloeosporioides and were further divided into asymptomatic and symptomatic samples. To characterize the global and local grayscale differences of the susceptible samples in different band images, the mean spectrum of the region of interest (ROI) of each band image was extracted as the global spectral features; moreover, the ROI in each band was segmented into disjointed local regions using the contrast limited adaptive histogram equalization and the Otsu thresholding algorithm, where the mean spectrum of the local regions were extracted as the local spectral features, respectively. The global and local spectral features were then concatenated into fused spectral features. Finally, the performance of the fused features was evaluated using support vector machine (SVM), k-NN and random forest (RF). The results showed that, compared with the conventional spectral feature-based methods, the proposed fused spectral features combined with SVM obtained the optimal results, where the average detection accuracy reached 91.97%. Furthermore, after applying feature selection using the successive projections algorithm (SPA), the resultant dimensionally-reduced fused spectral features also obtained acceptable results, with an average detection accuracy of 91.04%.

Why it matches plant phenotyping methods柑橘果実の病徴をハイパースペクトル画像と機械学習で検出する手法の開発・評価が中心であり、植物の疾病状態を直接推定しているため。

abstractAn early citrus anthracnose detection method using hyperspectral imaging and machine learning techniques is proposed.

Automated identification of citrus diseases in orchards using deep learning →

Citrus disease identification is vital to ensure the quality and quantity of production and minimise the damage in orchards. This study proposed a deep learning-based algorithm to perform automated identification on five common types of citrus diseases in orchards. The proposed algorithm consisted of a detection network to detect the citrus fruit in the complicated background and a classification network to classify them into the corresponding types. Several state-of-the-art network architectures were studied in terms of their object detection and classification performance, and they were evaluated on a dataset of 1524 images taken in field conditions from different orchards in distinct time intervals, scales, angles, and lighting conditions. Based on the experimental results, the algorithm eventually adopted an optimised YOLO-V4 model for detection and the EfficientNet model for classification, and the overall algorithm obtained the accuracy and F1 score of 0.890 and 0.872, respectively. In conclusion, the proposed algorithm is capable of automated citrus disease identification in orchards featuring high efficiency and precision.

Why it matches plant phenotyping methods柑橘果実画像から病害状態を自動識別する深層学習手法の開発と評価が中心であり、植物病害フェノタイプの画像ベース計測に該当する。

abstractThis study proposed a deep learning-based algorithm to perform automated identification on five common types of citrus diseases in orchards.

Sooty Mold Detection on Citrus Tree Canopy Using Deep Learning Algorithms. →

Sooty mold is a common disease found in citrus plants and is characterized by black fungi growth on fruits, leaves, and branches. This mold reduces the plant's ability to carry out photosynthesis. In small leaves, it is very difficult to detect sooty mold at the early stages. Deep learning-based image recognition techniques have the potential to identify and diagnose pest damage and diseases such as sooty mold. Recent studies used advanced and expensive hyperspectral or multispectral cameras attached to UAVs to examine the canopy of the plants and mid-range cameras to capture close-up infected leaf images. To bridge the gap on capturing canopy level images using affordable camera sensors, this study used a low-cost home surveillance camera to monitor and detect sooty mold infection on citrus canopy combined with deep learning algorithms. To overcome the challenges posed by varying light conditions, the main reason for using specialized cameras, images were collected at night, utilizing the camera's built-in night vision feature. A total of 4200 sliced night-captured images were used for training, 200 for validation, and 100 for testing, employed on the YOLOv5m, YOLOv7, and CenterNet models for comparison. The results showed that YOLOv7 was the most accurate in detecting sooty molds at night, with 74.4% mAP compared to YOLOv5m (72%) and CenterNet (70.3%). The models were also tested using preprocessed (unsliced) night images and day-captured sliced and unsliced images. The testing on preprocessed (unsliced) night images demonstrated the same trend as the training results, with YOLOv7 performing best compared to YOLOv5m and CenterNet. In contrast, testing on the day-captured images had underwhelming outcomes for both sliced and unsliced images. In general, YOLOv7 performed best in detecting sooty mold infections at night on citrus canopy and showed promising potential in real-time orchard disease monitoring and detection. Moreover, this study demonstrated that utilizing a cost-effective surveillance camera and deep learning algorithms can accurately detect sooty molds at night, enabling growers to effectively monitor and identify occurrences of the disease at the canopy level.

Why it matches plant phenotyping methods柑橘キャノピー上の病害状態を画像から検出する低コスト撮像・深層学習手法が中心で、複数モデルの比較検証も行っているため。

abstractthis study used a low-cost home surveillance camera to monitor and detect sooty mold infection on citrus canopy combined with deep learning algorithms.

A deep neural network with electronic nose for water stress prediction in Khasi Mandarin Orange plants →

Abstract Water stress is a significant environmental factor that hampers plant productivity and leads to various physiological and biological changes in plants. These include modifications in stomatal conductance and distribution, alteration of leaf water potential & turgor loss, altered chlorophyll content, and reduced cell expansion and growth. Additionally, water stress induces changes in the emission of volatile organic compounds across different parts of the plants. This study presents the development of an electronic nose (E-nose) system integrated with a deep neural network (DNN) to detect the presence and levels of water stress induced in Khasi Mandarin Orange plants. The proposed approach offers an alternative to conventional analytical methods that demand expensive and complex laboratory facilities. The investigation employs the leaf relative water content (RWC) estimation, a conventional technique, to evaluate water stress induction in the leaves of 20 plants collected over a span of 9 days after stopping irrigation. Supervised pattern recognition algorithms are trained using the results of RWC measurement, categorising leaves into non-stressed or one of four stress levels based on their water content. The dataset used for training and optimising the DNN model consists of 27 940 samples. The performance of the DNN model is compared to traditional machine learning methods, including linear and radial basis function support vector machines, k-nearest neighbours, decision tree, and random forest. From the results, it is seen that the optimised DNN model achieves the highest accuracy of 97.59% in comparison to other methods. Furthermore, the model is validated on an unseen dataset, exhibiting an accuracy of 97.32%. The proposed model holds the potential to enhance agricultural practices by enabling the detection and classification of water stress in crops, thereby aiding in water management improvements and increased productivity.

Why it matches plant phenotyping methods植物の水ストレスという生理状態を、電子鼻と深層学習で検出・分類する手法を開発し、従来法および未見データで性能検証しているため、植物フェノタイピング手法が中心である。

abstractThis study presents the development of an electronic nose (E-nose) system integrated with a deep neural network (DNN) to detect the presence and levels of water stress induced in Khasi Mandarin Orange plants.

A novel simulation model to predict photosynthetic active radiation interception in micro-irrigated citrus production orchards based on tree spacing, canopy geometry, and row orientation →

Diurnal and seasonal patterns of photosynthetically active radiation (PAR) intercepted by tree canopies in production orchards have an important role for eco-physiological processes that determine tree water use and productivity. Under well-watered conditions, PAR intercepted by trees is a reliable predictor of crop consumptive water use, which is regulated by tree spacing, tree row orientation, and by the tree canopy dimensions and geometry. Despite its advantages, only rare attempts have been made by the scientific and fruit production communities to develop canopy PAR interception simulators that rely on analytical principles to inform irrigation planning and management decisions. In this article, a novel model is presented that simulates PAR interception by citrus trees based on the canopy geometry and its shading patterns. The article also illustrates preliminary field testing and validation of the model with data collected in commercial citrus production orchards located in the San Joaquin Valley of California. The testing and validation could enable citrus growers and orchard managers to predict the net radiation for mature, micro-irrigated Navel orange and Page mandarin orchards grown with north–south and east–west tree row orientations and different tree densities. Results from the simulations and field validation showed that the canopy PAR light interception contributed 58.8 (± 0.02) and 71.1 (± 1.60) percent to the overall net radiation for the Navel orange orchards during the periods around summer solstice and autumnal equinox, respectively, whereas for the Page mandarin orchards, PAR light interception contributed 63.0 (±5.60) and 85.7 (±18.87) percent to the overall net radiation during the same time periods. Good correlations were found between daily values of PAR and net radiation for these citrus production orchards grown with the two alternative row orientations, which suggest that the proposed model could be used to inform irrigation planning and management decisions for different site-specific citrus orchard features and growing conditions.

Why it matches plant phenotyping methods柑橘樹冠によるPAR遮断を推定する新規シミュレーションモデルを開発し、圃場データで検証しており、植物樹冠の光環境状態の取得・推定手法が研究の中心である。

abstracta novel model is presented that simulates PAR interception by citrus trees based on the canopy geometry and its shading patterns.

Citrus Tree Canopy Segmentation of Orchard Spraying Robot Based on RGB-D Image and the Improved DeepLabv3+ →

The accurate and rapid acquisition of fruit tree canopy parameters is fundamental for achieving precision operations in orchard robotics, including accurate spraying and precise fertilization. In response to the issue of inaccurate citrus tree canopy segmentation in complex orchard backgrounds, this paper proposes an improved DeepLabv3+ model for fruit tree canopy segmentation, facilitating canopy parameter calculation. The model takes the RGB-D (Red, Green, Blue, Depth) image segmented canopy foreground as input, introducing Dilated Spatial Convolution in Atrous Spatial Pyramid Pooling to reduce computational load and integrating Convolutional Block Attention Module and Coordinate Attention for enhanced edge feature extraction. MobileNetV3-Small is utilized as the backbone network, making the model suitable for embedded platforms. A citrus tree canopy image dataset was collected from two orchards in distinct regions. Data from Orchard A was divided into training, validation, and test set A, while data from Orchard B was designated as test set B, collectively employed for model training and testing. The model achieves a detection speed of 32.69 FPS on Jetson Xavier NX, which is six times faster than the traditional DeepLabv3+. On test set A, the mIoU is 95.62%, and on test set B, the mIoU is 92.29%, showing a 1.12% improvement over the traditional DeepLabv3+. These results demonstrate the outstanding performance of the improved DeepLabv3+ model in segmenting fruit tree canopies under different conditions, thus enabling precise spraying by orchard spraying robots.

Why it matches plant phenotyping methodsRGB-D画像から果樹キャノピーを抽出し、キャノピー形状パラメータ算出に用いる改良セグメンテーション手法を開発・検証しており、植物形質取得が中心である。

abstractThe accurate and rapid acquisition of fruit tree canopy parameters is fundamental for achieving precision operations in orchard robotics

Citrus Huanglongbing detection based on polyphasic chlorophyll a fluorescence coupled with machine learning and model transfer in two citrus cultivars →

Citrus Huanglongbing (HLB) is regarded as the most severe disease threading the citrus industry. The basic aim of this study was to assess the changes of structure and functionality of the photosynthetic machinery induced by HLB and establish a discriminant model for rapid HLB detection using structural and functional features of photosynthesis system. Polyphasic chlorophyll a fluorescence transient was measured from healthy, asymptomatic and symptomatic HLB-infected leaves of two different cultivars namely (Navel orange and Satsuma). According to the polyphasic chlorophyll a fluorescence transient, HLB induced a positive L-band, K-band and J-band; a significant reduction in Fv/Fo, ΦEo, ψEo, ΦRo, ΦPo, PIₐbₛ, PIₜₒₜₐₗ, Fm, REo/RC and ETo/RC, and a significant increase in TRo/RC, Fo, ABS/RC and DIo/RC. The results suggested that the main disturbances of photosynthetic structure and function in HLB-infected leaves were associated with impairment of energetic connectivity of antennae in photosystem II (PSII), dysfunction of oxygen-evolving complex and inhibition of QA⁻ reoxidation. This phenomenon was similar in two different cultivars. Despite the fact that HLB infected symptomatic and magnesium deficient leaves can be often mistaken for each other by human eyes in the field condition, our results showed that some photosynthetic parameters could be considered as a potential proxy for distinguishing them from each other according to the results from principal component analysis (PCA). Moreover, the least squares support vector machine (LS-SVM) established based on the selected JIP-test parameters achieved overall detection accuracies of 95.0% for Navel orange and 96% for Satsuma using model transfer strategy. These results of OJIP records provided valuable information about the structure and function of photosynthetic apparatus in HLB infected leaves, and the photosynthetic fingerprints can be used for high-throughput HLB detection combining with advanced machine learning.

Why it matches plant phenotyping methodsHLB感染葉の光合成生理状態をクロロフィル蛍光で測定し、機械学習による迅速な病徴検出モデルを構築・評価しており、植物表現型の取得と抽出が中心的です。

abstractestablish a discriminant model for rapid HLB detection using structural and functional features of photosynthesis system

Citrus pose estimation from an RGB image for automated harvesting →

Automated fruit harvesting is promising research in the development of agricultural modernization. However, the complex and non-structural orchard environment is extremely challenging. In order to meet the needs of different end-effectors and to improve the success rate of automatic fruit harvesting, it is critical to perform fruit pose estimation before picking operations. In this study, a citrus pose estimation method through a single RGB image is introduced. The rotation of the citrus pose is defined as a vector that passes through the center of the fruit, which is perpendicular to the plane where the fruit navel point is located. Simply speaking, a multi-task learning model named FPENet is proposed to simultaneously locate the fruit navel point and predict the fruit rotation vector. And a hyperparameter is introduced in the loss function to achieve the simultaneous convergence of multiple tasks. In addition, this paper designs a 2D image annotation tool and constructs a citrus pose dataset, which contributes to model training and also the algorithm evaluation. In the experiment, we evaluate and analyze each module of the proposed network structure, and verify its performance on a harvesting robot. The experimental results show that the FPENet achieves an 88.92 AP score on fruit navel point detection, and 11.13° on the average error of the rotation vector. Over 90% of rotation vectors have an angular error of less than 22.5°. The harvesting success rate is 79.79%. This study offers a new idea for fruit pose estimation and provides the possibility and foundation for estimating fruit pose with a 2D image input.

Why it matches plant phenotyping methodsRGB画像から果実のへそ位置と回転ベクトル（果実姿勢）を推定する手法を開発し、データセットと注釈ツールも構築している。収穫対象の単なる検出・位置特定を超えて、再利用可能な果実器官の姿勢形質を抽出するため、中心的なフェノタイピング手法と判断する。

abstractIn this study, a citrus pose estimation method through a single RGB image is introduced.

Rapid Prediction of Nutrient Concentration in Citrus Leaves Using Vis-NIR Spectroscopy. →

The nutritional diagnosis of crops is carried out through costly foliar ionomic analysis in laboratories. However, spectroscopy is a sensing technique that could replace these destructive analyses for monitoring nutritional status. This work aimed to develop a calibration model to predict the foliar concentrations of macro and micronutrients in citrus plantations based on rapid non-destructive spectral measurements. To this end, 592 'Clementina de Nules' citrus leaves were collected during several months of growth. In these foliar samples, the spectral absorbance (430-1040 nm) was measured using a portable spectrometer, and the foliar ionomics was determined by emission spectrometry (ICP-OES) for macro and micronutrients, and the Kjeldahl method to quantify N. Models based on partial least squares regression (PLS-R) were calibrated to predict the content of macro and micronutrients in the leaves. The determination coefficients obtained in the model test were between 0.31 and 0.69, the highest values being found for P, K, and B (0.60, 0.63, and 0.69, respectively). Furthermore, the important P, K, and B wavelengths were evaluated using the weighted regression coefficients (BW) obtained from the PLS-R model. The results showed that the selected wavelengths were all in the visible region (430-750 nm) related to foliage pigments. The results indicate that this technique is promising for rapid and non-destructive foliar macro and micronutrient prediction.

Why it matches plant phenotyping methods携帯型Vis-NIR分光とPLS-Rによる柑橘葉の栄養濃度推定モデルを開発しており、植物形質取得・推定法が研究の中心である。

abstractThis work aimed to develop a calibration model to predict the foliar concentrations of macro and micronutrients in citrus plantations based on rapid non-destructive spectral measurements.

Machine Learning-based Citrus Plant Disease Detection and Management System using Computer Vision →

This article presents a novel method for identifying and managing diseases that affect citrus fruits by utilizing cutting-edge computer vision and machine learning techniques. The method is presented in this article. The proposed system utilizes a Convolutional Neural Network (CNN) model to extract features from images and classify them as healthy or diseased. Following this, a Random Forest algorithm is used to make the final prediction based on the extracted features. The system enables farmers to easily upload images of their citrus fruits from mobile or web-based platforms, enabling instant diagnosis of diseases and providing effective management plans to address the issue. The design of the proposed system is centered on enhancing the performance of the model in recognizing patterns that are associated with healthy and diseased fruits. In addition to this, it features an automatic management plan generation and alert system, both of which are designed to prompt farmers to take prompt actions for the prevention and control of diseases. The purpose of the proposed system is to deliver an instrument that is both effective and simple to use for the diagnosis and treatment of diseases, which will ultimately result in increased crop yields and increased profitability for citrus growers. The agricultural sector as a whole intends to gain significantly from the successful implementation of this system, which has the potential to bring about significant improvements.

Why it matches plant phenotyping methods柑橘果実画像から健全・罹病状態を推定するコンピュータビジョン手法が研究の中心であり、植物病害状態の画像ベース表現型計測に該当する。

abstractThis article presents a novel method for identifying and managing diseases that affect citrus fruits by utilizing cutting-edge computer vision and machine learning techniques.

Mandarin orange (Citrus reticulata Blanco cv. Batu 55) ripeness level prediction using combination reflectance-fluorescence spectroscopy. →

Optical characteristics of Mandarin Orange cv. Batu 55 of different maturity level has been obtained using reflectance (Vis-NIR) and fluorescence spectroscopy. Spectra features of both reflectance and fluorescence spectroscopy have been evaluated to develop a ripeness prediction model. Spectra dataset and reference measurements were subject to the partial least square regression (PLSR) analysis. The best prediction models were using reflectance spectroscopy data showing the coefficient of determination R 2 up to 0.89 and root mean square error (RMSE) of 2.71. On the other hand, it was found that fluorescence spectroscopy showed interesting spectra change in correlation with the accumulation of bluish and reddish fluorescence compounds in the lenticel spots on the fruit surface. The best prediction model using fluorescence spectroscopy data showed the R 2 of 0.88 and RMSE of 2.81. Besides that, it wa found that combining spectra of reflectance and fluorescence features could increase the R 2 of the partial least square regression (PLSR) model with Savitzky-Golay smoothing, up to 0.91 for brix-acid ratio prediction with RMSE 2.46. These results show the potential of the combined reflectance-fluorescence spectroscopy system for Mandarin ripeness assessment.

Why it matches plant phenotyping methodsマンダリン果実の熟度およびBrix-酸比という植物器官の状態を、反射・蛍光分光とPLSRで推定する測定システムを開発・評価しており、表現型取得法が研究の中心である。

abstractSpectra features of both reflectance and fluorescence spectroscopy have been evaluated to develop a ripeness prediction model.

Portable Mass Spectrometry Approach Combined with Machine Learning for Onsite Field Detection of Huanglongbing Disease. →

Huanglongbing (HLB) is one of the most serious citrus diseases in the world. Rapid, onsite, and accurate field detection of HLB is a challenging task in analytical science for a long time. Herein, we have developed a novel HLB detection method that combines headspace solid phase microextraction with portable gas chromatography-mass spectrometry (PGC-MS) approach for onsite field detection of volatile metabolites of citrus leaves. Detectability and characteristics of HLB-affected metabolites from leaves were validated, and the important biomarkers were verified by authentic compounds. A machine learning approach based on random forest algorithm is established to model the volatile metabolites from healthy, symptomatic, and asymptomatic citrus leaves. In this work, a total of 147 citrus leaf samples were analyzed. Analytical performances of this newly developed method were investigated by in-field detection of various volatile metabolites. Results demonstrated limits of detection and quantification of 0.04-0.12 and 0.17-0.44 ng/mL for different metabolites, respectively. Linear calibration curves of various metabolites were established over a concentration dynamic range of at least three orders ( R 2 > 0.96). Good reproducibility was obtained for intraday (3.0-17.5%, n = 6) and interday precision (8.7-18.2%, n = 7). This new HLB field detection method provides a rapid detection with 6 min for each sample via a simple optimized procedure, including onsite sampling, PGC-MS analysis, and data process and provides a high accuracy (93.3%) for simultaneous identification of healthy, symptomatic, and asymptomatic trees. These data support the use of this new method for reliable field detection of HLB. Furthermore, metabolic pathways of HLB-affected metabolites were also proposed. Overall, our results not only provide a rapid and onsite field HLB detection method but also provide valuable information for understanding metabolic change of HLB infection.

Why it matches plant phenotyping methods携帯型GC-MSと機械学習による柑橘葉の揮発性代謝物測定・HLB感染状態判定を中心に、検出性能と再現性を検証した植物表現型取得法である。

abstractwe have developed a novel HLB detection method that combines headspace solid phase microextraction with portable gas chromatography-mass spectrometry (PGC-MS) approach for onsite field detection of volatile metabolites of citrus leaves.

Classification of Citrus Huanglongbing Degree Based on CBAM-MobileNetV2 and Transfer Learning. →

Citrus has become a pivotal industry for the rapid development of agriculture and increasing farmers' incomes in the main production areas of southern China. Knowing how to diagnose and control citrus huanglongbing has always been a challenge for fruit farmers. To promptly recognize the diagnosis of citrus huanglongbing, a new classification model of citrus huanglongbing was established based on MobileNetV2 with a convolutional block attention module (CBAM-MobileNetV2) and transfer learning. First, the convolution features were extracted using convolution modules to capture high-level object-based information. Second, an attention module was utilized to capture interesting semantic information. Third, the convolution module and attention module were combined to fuse these two types of information. Last, a new fully connected layer and a softmax layer were established. The collected 751 citrus huanglongbing images, with sizes of 3648 × 2736, were divided into early, middle, and late leaf images with different disease degrees, and were enhanced to 6008 leaf images with sizes of 512 × 512, including 2360 early citrus huanglongbing images, 2024 middle citrus huanglongbing images, and 1624 late citrus huanglongbing images. In total, 80% and 20% of the collected citrus huanglongbing images were assigned to the training set and the test set, respectively. The effects of different transfer learning methods, different model training effects, and initial learning rates on model performance were analyzed. The results show that with the same model and initial learning rate, the transfer learning method of parameter fine tuning was obviously better than the transfer learning method of parameter freezing, and that the recognition accuracy of the test set improved by 1.02~13.6%. The recognition accuracy of the citrus huanglongbing image recognition model based on CBAM-MobileNetV2 and transfer learning was 98.75% at an initial learning rate of 0.001, and the loss value was 0.0748. The accuracy rates of the MobileNetV2, Xception, and InceptionV3 network models were 98.14%, 96.96%, and 97.55%, respectively, and the effect was not as significant as that of CBAM-MobileNetV2. Therefore, based on CBAM-MobileNetV2 and transfer learning, an image recognition model of citrus huanglongbing images with high recognition accuracy could be constructed.

Why it matches plant phenotyping methods柑橘葉の病害度という植物状態を画像から推定する深層学習モデルを開発・比較・評価しており、病害表現型の抽出手法が中心的である。

abstracta new classification model of citrus huanglongbing was established based on MobileNetV2 with a convolutional block attention module (CBAM-MobileNetV2) and transfer learning

Citrus disease classification with convolution neural network generated features and machine learning classifiers on hyperspectral image data →

Abstract has not been obtained from indexed metadata or an accessible article page.

Why it matches plant phenotyping methods柑橘の病害状態をハイパースペクトル画像から分類する機械学習手法が題名で明示され、植物病害表現型の取得・推定が中心である。

titleCitrus disease classification with convolution neural network generated features and machine learning classifiers on hyperspectral image data

Predicting and Visualizing Citrus Color Transformation Using a Deep Mask-Guided Generative Network. →

Citrus rind color is a good indicator of fruit development, and methods to monitor and predict color transformation therefore help the decisions of crop management practices and harvest schedules. This work presents the complete workflow to predict and visualize citrus color transformation in the orchard featuring high accuracy and fidelity. A total of 107 sample Navel oranges were observed during the color transformation period, resulting in a dataset containing 7,535 citrus images. A framework is proposed that integrates visual saliency into deep learning, and it consists of a segmentation network, a deep mask-guided generative network, and a loss network with manually designed loss functions. Moreover, the fusion of image features and temporal information enables one single model to predict the rind color at different time intervals, thus effectively shrinking the number of model parameters. The semantic segmentation network of the framework achieves the mean intersection over a union score of 0.9694, and the generative network obtains a peak signal-to-noise ratio of 30.01 and a mean local style loss score of 2.710, which indicate both high quality and similarity of the generated images and are also consistent with human perception. To ease the applications in the real world, the model is ported to an Android-based application for mobile devices. The methods can be readily expanded to other fruit crops with a color transformation period. The dataset and the source code are publicly available at GitHub.

Why it matches plant phenotyping methods柑橘果皮色を植物果実の発達状態として画像から推定・予測する手法を開発し、セグメンテーション、生成モデル、時系列情報を統合して技術性能を評価しているため、中心的なフェノタイピング手法研究である。

abstractA framework is proposed that integrates visual saliency into deep learning, and it consists of a segmentation network, a deep mask-guided generative network, and a loss network with manually designed loss functions.

SpectraNet-53: A deep residual learning architecture for predicting soluble solids content with VIS-NIR spectroscopy →

This work presents a new deep learning architecture, SpectraNet-53, for quantitative analysis of fruit spectra, optimized for predicting Soluble Solids Content (SSC, in °Brix). The novelty of this approach resides in being an architecture trainable on a very small dataset, while keeping a performance level on-par or above Partial Least Squares (PLS), a time-proven machine learning method in the field of spectroscopy. SpectraNet-53 performance is assessed by determining the SSC of 616 Citrus sinensi L. Osbeck 'Newhall' oranges, from two Algarve (Portugal) orchards, spanning two consecutive years, and under different edaphoclimatic conditions. This dataset consists of short-wave near-infrared spectroscopic (SW-NIRS) data, and was acquired with a portable spectrometer, in the visible to near infrared region, on-tree and without temperature equalization. SpectraNet-53 results are compared to a similar state-of-the-art architecture, DeepSpectra, as well as PLS, and thoroughly assessed on 15 internal validation sets (where the training and test data were sampled from the same orchard or year) and on 28 external validation sets (training/test data sampled from different orchards/years). SpectraNet-53 was able to achieve better performance than DeepSpectra and PLS in several metrics, and is especially robust to training overfit. For external validation results, on average, SpectraNet-53 was 3.1% better than PLS on RMSEP (1.16 vs. 1.20 °Brix), 11.6% better in SDR (1.22 vs. 1.10), and 28.0% better in R² (0.40 vs. 0.31).

Why it matches plant phenotyping methods果実スペクトルから糖度という植物器官形質を推定する深層学習アーキテクチャを開発し、既存手法と内部・外部検証で比較しており、表現型取得・推定法が中心である。

abstractThis work presents a new deep learning architecture, SpectraNet-53, for quantitative analysis of fruit spectra, optimized for predicting Soluble Solids Content (SSC, in °Brix).

System of counting green oranges directly from trees using artificial intelligence →

Abstract Agriculture is one of the most essential activities for humanity. Systems capable of automatically harvesting a crop using robots or performing a reasonable production estimate can reduce costs and increase production efficiency. With the advancement of computer vision, image processing methods are becoming increasingly viable in solving agricultural problems. Thus, this work aims to count green oranges directly from the trees through video footage filmed in line along a row of orange trees on the plantation. For the video image processing flow, a solution was proposed integrating the YOLOv4 network with object tracking algorithms. In order to compare the performance of the counting algorithm using the YOLOv4 network, an optimal object detector was simulated in which frame-by-frame corrected detections were used in which all oranges in all video frames were detected, and there were no erroneous detections. The use of YOLOv4 together with object detectors managed to reduce the number of double counting error and obtained a count close to the actual number of oranges visible in the video. The study also resulted in a database with an amount of 644 images with 43109 annotated oranges that can be used in future works.

Why it matches plant phenotyping methods樹上の果実数という植物器官の形質を、動画・YOLOv4・追跡アルゴリズムで自動抽出する手法を開発・評価しており、画像データセットも作成しているため、方法が中心的である。

abstractthis work aims to count green oranges directly from the trees through video footage

Classification of Appearance Quality of Red Grape Based on Transfer Learning of Convolution Neural Network →

Grapes are a globally popular fruit, with grape cultivation worldwide being second only to citrus. This article focuses on the low efficiency and accuracy of traditional manual grading of red grape external appearance and proposes a small-sample red grape external appearance grading model based on transfer learning with convolutional neural networks (CNNs). Initially, the CNN transfer learning method was used to transfer the pre-trained AlexNet, VGG16, GoogleNet, InceptionV3, and ResNet50 network models on the ImageNet image dataset to the red grape image grading task. By comparing the classification performance of the CNN models of these five different network depths with fine-tuning, ResNet50 with a learning rate of 0.001 and a loop number of 10 was determined to be the best feature extractor for red grape images. Moreover, given the small number of red grape image samples in this study, different convolutional layer features output by the ResNet50 feature extractor were analyzed layer by layer to determine the effect of deep features extracted by each convolutional layer on SVM classification performance. This analysis helped to obtain a ResNet50+SVM red grape external appearance grading model based on the optimal ResNet50 feature extraction strategy. Experimental data showed that the classification model constructed using the feature parameters extracted from the 10th node of the ResNet50 network achieved an accuracy rate of 95.08% for red grape grading. These research results provide a reference for the online grading of red grape clusters based on external appearance quality and have certain guiding significance for the quality and efficiency of grape industry circulation and production.

Why it matches plant phenotyping methodsブドウ果実房の外観品質を画像から分類・等級化するCNN転移学習手法が研究の中心であり、植物器官の形態的状態を抽出する実質的な画像ベース手法に該当する。

abstractproposes a small-sample red grape external appearance grading model based on transfer learning with convolutional neural networks (CNNs).

Design and Modeling of a Multi-camera-based Disease Detection Model →

Abstract A state-of-the-art approach for plant disease detection systems is discussed in this paper. Most proposed disease detection models in literature utilize single infeed cameras to capture the images of sample plant organs for classification. Single-input cameras might compromise the classification accuracy of these models depending on which plant organ is being used. Single input camera classification models have operated with high classification accuracy and efficiency with leaf samples because of their flat surface area nature, however, this is not always the case for fruit samples because of their general spherical or cylindrical nature such as oranges or bananas. The symptoms of a disease on the surface area of a sample fruit might not be distributed evenly, hence a single input camera sensor might miss the vital diseased part if the sample is orientated such that the diseased area is directly opposing to the line of sight of the camera sensor, which can consequently lead to an incorrect classification of that sample under evaluation. Hence, this study has proposed a multi-camera input fruit disease classification model aiming to solve this problem. Citrus orange fruits were used to demonstrate the capability of the proposed model to classify healthy and black rot-affected orange samples. A healthy sample and two black-rot-affected oranges, one with even and the other with uneven distribution of black rot symptoms, were put under evaluation of the proposed multi-camera input model and the classification accuracy was 100% when utilizing a deep learning Convolutional Neural Network classification algorithm.

Why it matches plant phenotyping methods植物果実の病徴を多視点カメラとCNNで分類する手法の提案・評価が中心であり、植物病害状態の画像ベース表現型計測に該当する。

abstractHence, this study has proposed a multi-camera input fruit disease classification model aiming to solve this problem.

Using EfficientNet and transfer learning for image-based diagnosis of nutrient deficiencies →

Early diagnosis of nutrient deficiencies can play a major role in avoiding significant agricultural losses and increasing the final yield while preserving the environment through efficient fertilizer usage. In this work, we study how well nutrient deficiency symptoms can be recognized in RGB images by using deep neural networks and transfer learning. Two different datasets, presenting real-world conditions, were used for this purpose. The first one was the Deep Nutrient Deficiency for Sugar Beet (DND-SB) dataset, which contains 5648 images of sugar beets presenting nitrogen (N), phosphorous (P), and potassium (K) deficiencies, the omission of liming (Ca) and full fertilization. The second one, collected on the field for this research and currently publicly available, was a dataset combining different orange tree images with iron (Fe), potasssium (K), magnesium (Mg), and manganese (Mn) deficiencies. Image classification via fine-tuning with EfficientNetB4, whose original weights came from a noisy student training on ImageNet, obtained the best performances on both datasets with 98.65% and 98.52% Top-1 accuracies. Additionally, the Grad-CAM++ analysis showed that the models were performing an accurate analysis of the most relevant part inside the images. Finally, the use of agricultural transfer learning did not report improvement in the performances.

Why it matches plant phenotyping methods植物の栄養欠乏症状をRGB画像から認識する深層学習手法を開発・比較評価し、複数データセットで性能検証しているため、表現型取得・判定手法が中心である。

abstractwe study how well nutrient deficiency symptoms can be recognized in RGB images by using deep neural networks and transfer learning.

A method of citrus epidermis defects detection based on an improved YOLOv5 →

Achieving intelligent detection of citrus epidermal defects after harvesting is of great significance to citrus quality and value assurance. The uneven illumination makes it challenging to detect citrus epidermis abnormalities with great accuracy. In order to improve the detection accuracy of citrus epidermal invisible defects, firstly, a dual-lamp image acquisition system is designed and used to complete the image acquisition of citrus fruit invisible defects. Secondly, the YOLOv5 model was optimized by integrating the attention mechanism CBAM and modifying the loss function as DIoU. Finally, the performance of the improved model was verified by comparison experiments and ablation experiments. According to the experimental results, mAP, Precision and Recall of the improved YOLOv5 model were 95.5%, 94.0% and 95.1%, respectively, which were 5.8%, 3.6% and 7.6% higher than those of YOLOv5x. Meanwhile, the average detection speed increased by 22.1 ms per pic. This indicates that the improved network being applied to citrus epidermal defects detection can achieve better performance. It can provide technical support for the intelligent detection and grading of postharvest citrus.

Why it matches plant phenotyping methods柑橘果実表皮の欠陥を画像取得と改良YOLOv5で検出する手法が研究の中心であり、比較・アブレーション実験による性能検証も行っている。

abstracta dual-lamp image acquisition system is designed and used to complete the image acquisition of citrus fruit invisible defects

Using Mobile Edge AI to Detect and Map Diseases in Citrus Orchards. →

Deep Learning models have presented promising results when applied to Agriculture 4.0. Among other applications, these models can be used in disease detection and fruit counting. Deep Learning models usually have many layers in the architecture and millions of parameters. This aspect hinders the use of Deep Learning on mobile devices as they require a large amount of processing power for inference. In addition, the lack of high-quality Internet connectivity in the field impedes the usage of cloud computing, pushing the processing towards edge devices. This work describes the proposal of an edge AI application to detect and map diseases in citrus orchards. The proposed system has low computational demand, enabling the use of low-footprint models for both detection and classification tasks. We initially compared AI algorithms to detect fruits on trees. Specifically, we analyzed and compared YOLO and Faster R-CNN. Then, we studied lean AI models to perform the classification task. In this context, we tested and compared the performance of MobileNetV2, EfficientNetV2-B0, and NASNet-Mobile. In the detection task, YOLO and Faster R-CNN had similar AI performance metrics, but YOLO was significantly faster. In the image classification task, MobileNetMobileV2 and EfficientNetV2-B0 obtained an accuracy of 100%, while NASNet-Mobile had a 98% performance. As for the timing performance, MobileNetV2 and EfficientNetV2-B0 were the best candidates, while NASNet-Mobile was significantly worse. Furthermore, MobileNetV2 had a 10% better performance than EfficientNetV2-B0. Finally, we provide a method to evaluate the results from these algorithms towards describing the disease spread using statistical parametric models and a genetic algorithm to perform the parameters' regression. With these results, we validated the proposed pipeline, enabling the usage of adequate AI models to develop a mobile edge AI solution.

Why it matches plant phenotyping methods柑橘の病害を画像から検出・分類し、病害拡散を推定するエッジAIパイプラインの開発と比較検証が研究の中心であり、植物の病害状態を測定するフェノタイピング手法に該当する。

abstractThis work describes the proposal of an edge AI application to detect and map diseases in citrus orchards.

Estimating stomatal conductance of citrus under water stress based on multispectral imagery and machine learning methods. →

Introduction Canopy stomatal conductance (Sc) indicates the strength of photosynthesis and transpiration of plants. In addition, Sc is a physiological indicator that is widely employed to detect crop water stress. Unfortunately, existing methods for measuring canopy Sc are time-consuming, laborious, and poorly representative. Methods To solve these problems, in this study, we combined multispectral vegetation index (VI) and texture features to predict the Sc values and used citrus trees in the fruit growth period as the research object. To achieve this, VI and texture feature data of the experimental area were obtained using a multispectral camera. The H (Hue), S (Saturation) and V (Value) segmentation algorithm and the determined threshold of VI were used to obtain the canopy area images, and the accuracy of the extraction results was evaluated. Subsequently, the gray level co-occurrence matrix (GLCM) was used to calculate the eight texture features of the image, and then the full subset filter was used to obtain the sensitive image texture features and VI. Support vector regression, random forest regression, and k-nearest neighbor regression (KNR) Sc prediction models were constructed, which were based on single and combined variables. Results The analysis revealed the following: 1) the accuracy of the HSV segmentation algorithm was the highest, achieving more than 80%. The accuracy of the VI threshold algorithm using excess green was approximately 80%, which achieved accurate segmentation. 2) The citrus tree photosynthetic parameters were all affected by different water supply treatments. The greater the degree of water stress, the lower the net photosynthetic rate (Pn), transpiration rate (Tr), and Sc of the leaves. 3) In the three Sc prediction models, The KNR model, which was constructed by combining image texture features and VI had the optimum prediction effect (training set: R 2 = 0.91076, RMSE = 0.00070; validation set; R 2 = 0.77937, RMSE = 0.00165). Compared with the KNR model, which was only based on VI or image texture features, the R 2 of the validation set of the KNR model based on combined variables was improved respectively by 6.97% and 28.42%. Discussion This study provides a reference for large-scale remote sensing monitoring of citrus Sc by multispectral technology. Moreover, it can be used to monitor the dynamic changes of Sc and provide a new technique for gaining a better understanding of the growth status and water stress of citrus crops.

Why it matches plant phenotyping methodsマルチスペクトル画像から気孔コンダクタンスという植物生理形質を抽出・予測する画像処理および機械学習手法が研究の中心であり、分割精度と予測性能も検証している。

abstractwe combined multispectral vegetation index (VI) and texture features to predict the Sc values

Characteristics of photosynthesis and vertical canopy architecture of citrus trees under two labor-saving cultivation modes using unmanned aerial vehicle (UAV)-based LiDAR data in citrus orchards. →

Analyzing and comparing the effects of labor-saving cultivation modes on photosynthesis, as well as studying their vertical canopy architecture, can improve the tree structure of high-quality and high-yield citrus and selection of labor-saving cultivation modes. The photosynthesis of 1080 leaves of two labor-saving cultivation modes (wide-row and narrow-plant mode and fenced mode) comparing with the traditional mode were measured, and nitrogen content of all leaves and photosynthetic nitrogen use efficiency (PNUE) were determined. Unmanned aerial vehicle (UAV)-based light detection and ranging (LiDAR) data were used to assess the vertical architecture of three citrus cultivation modes. Results showed that for the wide-row and narrow-plant and traditional modes leaf photosynthetic CO 2 assimilation rate, stomatal conductance, and transpiration rate of the upper layer were significantly higher than those of the middle layer, and values of the middle layer were markedly higher than those of the lower layer. In the fenced mode, a significant difference in photosynthetic factors between the upper and middle layers was not observed. A vertical canopy distribution had a more significant effect on PNUE in the traditional mode. Leaves in the fenced mode had distinct photosynthetic advantages and higher PNUE. UAV-based LiDAR data effectively revealed the differences in the vertical canopy architecture of citrus trees by enabling calculating the density and height percentile of the LiDAR point cloud. The point cloud densities of three cultivation modes were significantly different for all LiDAR density slices, especially at higher canopy heights. The labor-saving modes, particularly the fenced mode, had significantly higher height percentile data.

Why it matches plant phenotyping methodsUAV-LiDARによる点群密度・高さパーセンタイルの算出を用いて、柑橘樹の垂直樹冠構造を定量評価しており、植物形態表現型の取得・抽出が研究の主要部分に含まれる。

abstractUnmanned aerial vehicle (UAV)-based light detection and ranging (LiDAR) data were used to assess the vertical architecture of three citrus cultivation modes.

Performance evaluation of plant leaf disease detection using deep learning models →

Plant diseases have a serious impact on production, and hence they must be detected and recognised at early stages. Smart firming using deep learning can automatically identify infected crops. We provide extremely effective convolution neural network (CNN) designs for the identification of leaf diseases in this research strategy. For the training and testing phases of this study, a database of potato leaves is produced. To classify the disease from the input photos of the supported training dataset, we employed CNN to extract its characteristics. 1700 photos of potato leaves were used for model training, and then about 600 images were used for testing. To identify citrus diseases, Convolutional Neural Networks, Deep Learning, base learning, and transfer learning were applied. Results from training, testing, and experiments indicate that the suggested architecture has outperformed other current models in terms of ResNet model accuracy, achieving a score of 99.62%.

Why it matches plant phenotyping methods葉画像から植物病害を推定するCNN手法を開発・比較評価しており、植物の病害状態の取得が研究の中心である。

titlePerformance evaluation of plant leaf disease detection using deep learning models

Intelligent detection of citrus fruit pests using machine vision system and convolutional neural network through transfer learning technique. →

Plant pests and diseases play a significant role in reducing the quality of agricultural products. As one of the most important plant pathogens, pests like Mediterranean fruit fly cause significant damage to crops and thus annually farmers face a lot of loss in their products. Therefore, the use of modern and non-destructive methods such as machine vision systems and deep learning for early detection of pests in agricultural products is of particular importance. In this study, citrus fruit images were taken in three stages: 1) before pest infestation, 2) beginning of fruit infestation, and 3) eight days after the second stage, in natural light conditions (7000-11,000 lux). A total of 1519 images were prepared for all classes. To classify the images, 70% of the images were used for the network training stage, 10% and 20% of the images were used for the validation and testing stages. Four pre-trained CNN models, namely ResNet-50, GoogleNet, VGG-16 and AlexNet as well as the SGDm, RMSProp and Adam optimization algorithms were used to identify and classify healthy fruit and fruit infected with the Mediterranean fly. The results of evaluating the models in the pest outbreak stage showed that the VGG-16 model with the help of SGDm algorithm had the best efficiency with the highest detection accuracy and F1 of 98.33% and 98.36%, respectively. The evaluation of the third stage showed that the AlexNet model with the help of SGDm algorithm had the best result with the highest detection accuracy and F1 of 99.33% and 99.34%, respectively. AlexNet model using SGDm optimization algorithm had the shortest network training time (323 s). The results of this study showed that convolutional neural network method and machine vision system can be effective in controlling and managing pests in orchards and other agricultural products.

Why it matches plant phenotyping methods柑橘果実の画像から害虫感染状態をCNNで検出・分類する画像ベースの植物状態推定法が研究の中心であり、性能評価も実施している。

abstractthe use of modern and non-destructive methods such as machine vision systems and deep learning for early detection of pests in agricultural products is of particular importance

Custom-Developed Reflection-Transmission Integrated Vision System for Rapid Detection of Huanglongbing Based on the Features of Blotchy Mottled Texture and Starch Accumulation in Leaves. →

Huanglongbing (HLB) is a highly contagious and devastating citrus disease that causes huge economic losses to the citrus industry. Because it cannot be cured, timely detection of the HLB infection status of plants and removal of diseased trees are effective ways to reduce losses. However, complex HLB symptoms, such as single HLB-symptomatic or zinc deficiency + HLB-positive, cannot be identified by a single reflection imaging method at present. In this study, a vision system with an integrated reflection-transmission image acquisition module, human-computer interaction module, and power supply module was developed for rapid HLB detection in the field. In reflection imaging mode, 660 nm polarized light was used as the illumination source to enhance the contrast of the HLB symptoms in the images based on the differences in the absorption of narrow-band light by the components within the leaves. In transmission imaging mode, polarization images were obtained in four directions, and the polarization angle images were calculated using the Stokes vector to detect the optical activity of starch. A step-by-step classification model with four steps was used for the identification of six classes of samples (healthy, HLB-symptomatic, zinc deficiency, zinc deficiency + HLB-positive, magnesium deficiency, and boron deficiency). The results showed that the model had an accuracy of 96.92% for the full category of samples and 98.08% for the identification of multiple types of HLB (HLB-symptomatic and zinc deficiency + HLB-positive). In addition, the classification model had good recognition of zinc deficiency and zinc deficiency + HLB-positive samples, at 92.86%.

Why it matches plant phenotyping methods植物葉の病徴・デンプン蓄積を反射／透過画像で取得し、HLB感染状態を推定する統合ビジョンシステムを開発・評価しており、植物フェノタイピング手法が研究の中心である。

abstracta vision system with an integrated reflection-transmission image acquisition module, human-computer interaction module, and power supply module was developed for rapid HLB detection in the field.

Flexible strain sensor with a hat-shaped structure for in situ measurement of 3D deformation →

Flexible strain sensors that are currently available are mainly used in human motion recognition and medical health detection applications, and there is still an urgent need for sensors to realize real-time monitoring of the 3D deformation of industrial and agricultural products. In this work, a flexible strain sensor with a hat-shaped structure was fabricated using a molding technique to perform in situ measurement of 3D deformation. An algorithm for resistance change detection and linear calibration equations were proposed to enable analysis of the deformation data and calculation of local shape changes. The sensor was applied to monitor the growth deformation of a kumquat fruit, and the results were highly consistent with the algorithm. The proposed technique has great potential for application to 3D deformation detection of flexible objects.

Why it matches plant phenotyping methods柔軟ひずみセンサーと抵抗変化解析・校正式を開発し、カンキツ果実の成長変形という植物器官の形質を実測・検証しており、フェノタイピング手法が中心である。

abstractAn algorithm for resistance change detection and linear calibration equations were proposed to enable analysis of the deformation data and calculation of local shape changes.

PFDI: A Precise Fruit disease Identification Model based on Context Data Fusion with Faster-CNN in Edge Computing Environment →

Abstract Fruits have a significant impact on everyday living i.e., citrus fruits. Numerous fruits have a solid nutritious value and are packed with multivitamins and trace components. Citrus fruits are delicate, so they are susceptible to many diseases and infections. Many researchers have suggested various deep learning and machine learning based fruit disease detection and classification models. In this research we are presenting precise fruit disease identification (PFDI) model based on context data fusion with Faster-CNN in edge computing environment. The goal is to develop a precise, efficient, and trustable fruit disease detection model, which is a critical component of an autonomous food production in robotic edge platform. This research examines and explores four different diseases of citrus fruits using CNN deep learning models to be adopted as edge computing solution. Identification of citrus diseases namely cankers black spot, greening, scab, melanose, and healthy citrus fruits are implemented using the proposed sequential model without pruning, with pruning having different sparsity level followed by post quantization. Through transfer learning method, we optimize this model for the assignment of fruit disease detection employing visuals from two patterns: Near-infrared (NIFR) and RGB. For integrating multi-model (NIFR and RGB) facts, early and late data fusion techniques are evaluated. The accuracy obtained from the proposed model for the canker disease is 97%,scab 95%, melanose 99% ,Greening 97%,Black spot 97% and for healthy 97%. In this paper we compared and evaluated the results of proposed model with the sparsity levels of 50–80%, 60–90%, 70–90%, 80–90% pruning and also obtained the results of post-quantization on each level. The results show that the size of the model with 60–90% pruning can be counteracted to the 47.64 of the baseline model without significant loss of accuracy. Moreover, post quantization can further reduces the of 60–90% pruning from 28.16 to 8.72. In addition to enhanced precision, the above initiative is much faster to implement for new fruits diseases because it needs bounding box annotation (BBA) instead of pixel-level annotation (PLA).

Why it matches plant phenotyping methods柑橘の病徴を画像から検出・分類する深層学習手法の開発と、NIR/RGB融合、剪定、量子化、精度評価を扱っており、植物状態の取得・推定が中心である。

abstractwe are presenting precise fruit disease identification (PFDI) model based on context data fusion with Faster-CNN in edge computing environment.

Towards sweetness classification of orange cultivars using short-wave NIR spectroscopy. →

The global orange industry constantly faces new technical challenges to meet consumer demands for quality fruits. Instead of traditional subjective fruit quality assessment methods, the interest in the horticulture industry has increased in objective, quantitative, and non-destructive assessment methods. Oranges have a thick peel which makes their non-destructive quality assessment challenging. This paper evaluates the potential of short-wave NIR spectroscopy and direct sweetness classification approach for Pakistani cultivars of orange, i.e., Red-Blood, Mosambi, and Succari. The correlation between quality indices, i.e., Brix, titratable acidity (TA), Brix: TA and BrimA (Brix minus acids), sensory assessment of the fruit, and short-wave NIR spectra, is analysed. Mix cultivar oranges are classified as sweet, mixed, and acidic based on short-wave NIR spectra. Short-wave NIR spectral data were obtained using the industry standard F-750 fruit quality meter (310-1100 nm). Reference Brix and TA measurements were taken using standard destructive testing methods. Reference taste labels i.e., sweet, mix, and acidic, were acquired through sensory evaluation of samples. For indirect fruit classification, partial least squares regression models were developed for Brix, TA, Brix: TA, and BrimA estimation with a correlation coefficient of 0.57, 0.73, 0.66, and 0.55, respectively, on independent test data. The ensemble classifier achieved 81.03% accuracy for three classes (sweet, mixed, and acidic) classification on independent test data for direct fruit classification. A good correlation between NIR spectra and sensory assessment is observed as compared to quality indices. A direct classification approach is more suitable for a machine-learning-based orange sweetness classification using NIR spectroscopy than the estimation of quality indices.

Why it matches plant phenotyping methodsオレンジ果実の糖度・酸度・甘味を非破壊NIR分光と機械学習で推定・分類する方法が研究の中心であり、果実品質という植物形質を直接評価している。

abstractThis paper evaluates the potential of short-wave NIR spectroscopy and direct sweetness classification approach for Pakistani cultivars of orange

Hyperspectral Imaging with Machine Learning Approaches for Assessing Soluble Solids Content of Tribute Citru. →

Tribute Citru is a natural citrus hybrid with plenty of vitamins and nutrients. Fruits' soluble solids content (SSC) is a critical quality index. This study used hyperspectral imaging at two spectral ranges (400-1000 nm and 900-1700 nm) to determine SSC in Tribute Citru. Partial least squares regression (PLSR) and support vector regression (SVR) models were established in order to determine SSC using the spectral information of the calyx and blossom ends. The average spectra of both ends as well as their fusion was studied. The successive projections algorithm (SPA) and the correlation coefficient analysis (CCA) were used to examine the differences in characteristic wavelengths between the two ends. Most models achieved performances with the correlation coefficient of the training, validation, and testing sets over 0.6. Results showed that differences in the performances among the models using the one-sided and two-sided spectral information. No particular regulation could be found for the differences in model performances and characteristic wavelengths. The results illustrated that the sampling side was an influencing factor but not the determinant factor for SSC determination. These results would help with the development of real-world applications for citrus quality inspection without concerning the sampling sides and the spectral ranges.

Why it matches plant phenotyping methods柑橘果実の可溶性固形分という植物器官の形質を、ハイパースペクトル画像と機械学習で推定する方法を開発・比較しており、表現型取得・推定が中心である。

abstractThis study used hyperspectral imaging at two spectral ranges (400-1000 nm and 900-1700 nm) to determine SSC in Tribute Citru.

Automatic Fruit Disease Classification using Machine Learning Strategies for Agriculture Farming →

Fruit protection is critical in the agricultural industry in the context of the global economy. The general public has recently learned that various diseases are wreaking havoc on fruit supplies. Agriculture economies all over the world are failing as a result of this. Computerized automatic methods for assessing the quality of fresh and rotting apples can relieve the burden of manually researching various apple fruit varieties. This study presents a novel method for comparing apple varieties based on fruit quality. The technique employs principal component analysis (PCA) early on to collect relevant characteristics. Additionally, the statistical, textual, and geometrical features are also extracted. The model is first tested by classifying apples from the Fruits-360 dataset as fresh or spoiled. Furthermore, we classified four different types of fresh and rotting fruits during the training and testing phases of the classification procedure (apple, avocado, banana, and orange). In addition, the k-NN algorithm, the linear support vector machine (LSVM), the kernel support vector machine (KSVM), and decision trees (DT) are classifiers to categorize the fruits according to quality. After evaluating the models’ performance, they are retrained using only the first two principal components. The results of using SVM and DT models for quality evaluation have been discovered to be more encouraging and comparable to those obtained using state-of-the-art methods.

Why it matches plant phenotyping methods果実画像から鮮度・腐敗状態を抽出・分類する機械学習手法が研究の中心であり、植物器官の状態を直接評価しているため採用。

abstractComputerized automatic methods for assessing the quality of fresh and rotting apples can relieve the burden of manually researching various apple fruit varieties.

Citrus Tree Crown Segmentation of Orchard Spraying Robot Based on RGB-D Image and Improved Mask R-CNN →

Orchard spraying robots must visually obtain citrus tree crown growth information to meet the variable growth-stage-based spraying requirements. However, the complex environments and growth characteristics of fruit trees affect the accuracy of crown segmentation. Therefore, we propose a feature-map-based squeeze-and-excitation UNet++ (MSEU) region-based convolutional neural network (R-CNN) citrus tree crown segmentation method that intakes red–green–blue-depth (RGB-D) images that are pixel aligned and visual distance-adjusted to eliminate noise. Our MSEU R-CNN achieves accurate crown segmentation using squeeze-and-excitation (SE) and UNet++. To fully fuse the feature map information, the SE block correlates image features and recalibrates their channel weights, and the UNet++ semantic segmentation branch replaces the original mask structure to maximize the interconnectivity between feature layers, achieving a near-real time detection speed of 5 fps. Its bounding box (bbox) and segmentation (seg) AP50 scores are 96.6 and 96.2%, respectively, and the bbox average recall and F1-score are 73.0 and 69.4%, which are 3.4, 2.4, 4.9, and 3.5% higher than the original model, respectively. Compared with bbox instant segmentation (BoxInst) and conditional convolutional frameworks (CondInst), the MSEU R-CNN provides better seg accuracy and speed than the previous-best Mask R-CNN. These results provide the means to accurately employ autonomous spraying robots.

Why it matches plant phenotyping methodsRGB-D画像から柑橘樹冠を抽出する画像解析手法を開発・評価しており、樹冠という植物形態・生育状態の取得が中心であるため、ロボット用途でも植物フェノタイピング手法に該当する。

abstractwe propose a feature-map-based squeeze-and-excitation UNet++ (MSEU) region-based convolutional neural network (R-CNN) citrus tree crown segmentation method that intakes red–green–blue-depth (RGB-D) images

An automatic identification system for citrus greening disease (Huanglongbing) using a YOLO convolutional neural network →

Huanglongbing (HLB), or citrus greening disease, has complex and variable symptoms, making its diagnosis almost entirely reliant on subjective experience, which results in a low diagnosis efficiency. To overcome this problem, we constructed and validated a deep learning (DL)-based method for detecting citrus HLB using YOLOv5l from digital images. Three models (Yolov5l-HLB1, Yolov5l-HLB2, and Yolov5l-HLB3) were developed using images of healthy and symptomatic citrus leaves acquired under a range of imaging conditions. The micro F1-scores of the Yolov5l-HLB2 model (85.19%) recognising five HLB symptoms (blotchy mottling, "red-nose" fruits, zinc-deficiency, vein yellowing, and uniform yellowing) in the images were higher than those of the other two models. The generalisation performance of Yolov5l-HLB2 was tested using test set images acquired under two photographic conditions (conditions B and C) that were different from that of the model training set condition (condition A). The results suggested that this model performed well at recognising the five HLB symptom images acquired under both conditions B and C, and yielded a micro F1-score of 84.64% and 85.84%, respectively. In addition, the detection performance of the Yolov5l-HLB2 model was better for experienced users than for inexperienced users. The PCR-positive rate of Candidatus Liberibacter asiaticus (CLas) detection (the causative pathogen for HLB) in the samples with five HLB symptoms as classified using the Yolov5l-HLB2 model was also compared with manual classification by experts. This indicated that the model can be employed as a preliminary screening tool before the collection of field samples for subsequent PCR testing. We also developed the 'HLBdetector' app using the Yolov5l-HLB2 model, which allows farmers to complete HLB detection in seconds with only a mobile phone terminal and without expert guidance. Overall, we successfully constructed a reliable automatic HLB identification model and developed the user-friendly 'HLBdetector' app, facilitating the prevention and timely control of HLB transmission in citrus orchards.

Why it matches plant phenotyping methods柑橘葉や果実のHLB症状という植物状態を画像から検出する深層学習法を開発・検証し、アプリ化しており、表現型取得が研究の中心である。

abstractwe constructed and validated a deep learning (DL)-based method for detecting citrus HLB using YOLOv5l from digital images.

Citrus disease detection using convolution neural network generated features and Softmax classifier on hyperspectral image data →

Identification and segregation of citrus fruit with diseases and peel blemishes are required to preserve market value. Previously developed machine vision approaches could only distinguish cankerous from non-cankerous citrus, while this research focused on detecting eight different peel conditions on citrus fruit using hyperspectral (HSI) imagery and an AI-based classification algorithm. The objectives of this paper were: (i) selecting the five most discriminating bands among 92 using PCA, (ii) training and testing a custom convolution neural network (CNN) model for classification with the selected bands, and (iii) comparing the CNN's performance using 5 PCA bands compared to five randomly selected bands. A hyperspectral imaging system from earlier work was used to acquire reflectance images in the spectral region from 450 to 930 nm (92 spectral bands). Ruby Red grapefruits with normal, cankerous, and 5 other common peel diseases including greasy spot, insect damage, melanose, scab, and wind scar were tested. A novel CNN based on the VGG-16 architecture was developed for feature extraction, and SoftMax for classification. The PCA-based bands were found to be 666.15, 697.54, 702.77, 849.24 and 917.25 nm, which resulted in an average accuracy, sensitivity, and specificity of 99.84%, 99.84% and 99.98% respectively. However, 10 trials of five randomly selected bands resulted in only a slightly lower performance, with accuracy, sensitivity, and specificity of 98.87%, 98.43% and 99.88%, respectively. These results demonstrate that an AI-based algorithm can successfully classify eight different peel conditions. The findings reported herein can be used as a precursor to develop a machine vision-based, real-time peel condition classification system for citrus processing.

Why it matches plant phenotyping methods柑橘果実の病害・果皮状態をハイパースペクトル画像とCNNで分類する画像ベースの植物状態フェノタイピング手法を開発・評価しており、手法が研究の中心である。

abstractthis research focused on detecting eight different peel conditions on citrus fruit using hyperspectral (HSI) imagery and an AI-based classification algorithm.

Using infrared thermal imaging technology to estimate the transpiration rate of citrus trees and evaluate plant water status →

Abstract has not been obtained from indexed metadata or an accessible article page.

Why it matches plant phenotyping methods赤外線熱画像を用いて柑橘樹の蒸散速度と水分状態という植物生理形質を推定する方法が題名で明示されており、フェノタイピング手法の適用が中心です。

titleUsing infrared thermal imaging technology to estimate the transpiration rate of citrus trees and evaluate plant water status

Citrus green fruit detection via improved feature network extraction. →

Introduction It is crucial to accurately determine the green fruit stage of citrus and formulate detailed fruit conservation and flower thinning plans to increase the yield of citrus. However, the color of citrus green fruits is similar to the background, which results in poor segmentation accuracy. At present, when deep learning and other technologies are applied in agriculture for crop yield estimation and picking tasks, the accuracy of recognition reaches 88%, and the area enclosed by the PR curve and the coordinate axis reaches 0.95, which basically meets the application requirements.To solve these problems, this study proposes a citrus green fruit detection method that is based on improved Mask-RCNN (Mask-Region Convolutional Neural Network) feature network extraction. Methods First, the backbone networks are able to integrate low, medium and high level features and then perform end-to-end classification. They have excellent feature extraction capability for image classification tasks. Deep and shallow feature fusion is used to fuse the ResNet(Residual network) in the Mask-RCNN network. This strategy involves assembling multiple identical backbones using composite connections between adjacent backbones to form a more powerful backbone. This is helpful for increasing the amount of feature information that is extracted at each stage in the backbone network. Second, in neural networks, the feature map contains the feature information of the image, and the number of channels is positively related to the number of feature maps. The more channels, the more convolutional layers are needed, and the more computation is required, so a combined connection block is introduced to reduce the number of channels and improve the model accuracy. To test the method, a visual image dataset of citrus green fruits is collected and established through multisource channels such as handheld camera shooting and cloud platform acquisition. The performance of the improved citrus green fruit detection technology is compared with those of other detection methods on our dataset. Results The results show that compared with Mask-RCNN model, the average detection accuracy of the improved Mask-RCNN model is 95.36%, increased by 1.42%, and the area surrounded by precision-recall curve and coordinate axis is 0.9673, increased by 0.3%. Discussion This research is meaningful for reducing the effect of the image background on the detection accuracy and can provide a constructive reference for the intelligent production of citrus.

Why it matches plant phenotyping methods柑橘青果の画像検出手法を改良・評価し、果実という植物器官の状態を直接推定しているため、方法が研究の中心である。

abstractthis study proposes a citrus green fruit detection method that is based on improved Mask-RCNN (Mask-Region Convolutional Neural Network) feature network extraction.

Terahertz spectroscopic monitoring and analysis of citrus leaf water status under low temperature stress. →

Low temperature stress, in the form of chilling and freezing, is one of the major environmental factors impacting on citrus yield, which changes plant's water state and results in the crops' sub-health or injury. The innovative terahertz (THz) spectroscopy and imaging based sensing technology has been shown to be a suitable tool for plant leaf water status determination, due to THz radiation's innate sensitivity to hydrogen bond vibration in aqueous solutions, which is usually related to plant phenotype change. We demonstrate experimentally that the THz absorption coefficient of leaf could be used for distinguishing plant's physiological stress status, exhibiting clear decreasing or increasing trend under chilling or freezing stress respectively. The underlying rationale might be that membrane damage shows a diverse pattern, changing the intra- or extra-cellular liquid environments, likely being linked to the various THz spectral characteristics. There were different adaptations in leaf morphology, leading to different leaf density, which in turn affects the water volume fraction. Moreover, different patterns of the dynamic equilibrium state of free water and bound water under chilling and freezing treatment were revealed by THz spectroscopy. Here, THz spectroscopic monitoring has shown unique potential for judging citrus's low temperature stress state through bio-water detection and discrimination.

Why it matches plant phenotyping methodsTHz分光・イメージングを用いて柑橘葉の水分状態と低温ストレス状態を判定するセンシング手法が研究の中心であり、植物の生理状態を直接推定している。

abstractThe innovative terahertz (THz) spectroscopy and imaging based sensing technology has been shown to be a suitable tool for plant leaf water status determination

A novel approach for measuring membrane permeability for organic compounds via surface plasmon resonance detection. →

Well-known methods for measuring permeability of membranes include static or flow diffusion chambers. When studying the effects of organic compounds on plants, the use of such model systems allows to investigate xenobiotic behavior at the cuticular barrier level and obtain an understanding of the initial penetration processes of these substances into plant leaves. However, the use of diffusion chambers has disadvantages, including being time-consuming, requiring sampling, or a sufficiently large membrane area, which cannot be obtained from all types of plants. Therefore, we propose a new method based on surface plasmon resonance imaging (SPRi) to enable rapid membrane permeability evaluation. This study presents the methodology for measuring permeability of isolated cuticles for organic compounds via surface plasmon resonance detection, where the selected model analyte was the widely used pesticide metazachlor. Experiments were performed on the cuticles of Ficus elastica, Citrus pyriformis, and an artificial PES membrane, which is used in passive samplers for the detection of xenobiotics in water and soils. The average permeability for metazachlor was 5.23 × 10 -14 m 2 s -1 for C. pyriformis, 1.34 × 10 -13 m 2 s -1 for F. elastica, and 7.74 × 10 -12 m 2 s -1 for the PES membrane. We confirmed that the combination of a flow-through diffusion cell and real-time optical detection of transposed molecules represents a promising method for determining the permeability of membranes to xenobiotics occurring in the environment. This is necessary for determining a pesticide dosage in agriculture, selecting suitable membranes for passive samplers in analytics, testing membranes for water treatment, or studying material use of impregnated membranes.

Why it matches plant phenotyping methods植物葉の単離クチクラにおける有機化合物透過性をSPRで測定する新手法を開発しており、植物組織の生理的特性の取得が研究の中心である。

abstractTherefore, we propose a new method based on surface plasmon resonance imaging (SPRi) to enable rapid membrane permeability evaluation.

P2OP—Plant Pathology on Palms: A deep learning-based mobile solution for in-field plant disease detection →

Plant diseases are one of the dominant factors that threaten sustainable agriculture, leading to economic losses. Developing an accurate mobile-based plant disease detection methodology is important for enabling rapid identification of emerging diseases directly from the farms. The deep learning methods have limited usage in mobile-based applications as they require larger memory and processing power to operate directly on smartphones or internet connectivity when used with a client–server computing model. To address this challenge, we propose a mobile-based lightweight deep learning-based model, which requires only a small footprint and processing power while maintaining higher detection accuracy. With around 0.088 billion multiply–accumulation operations, 0.26 million parameters, and 1 MB storage space, this framework achieved 97%, 97.1% and 96.4% accuracies on apple, citrus and tomato leaves datasets, respectively. One of our tiny models achieved 93.33% accuracy on a custom sourced in-the-wild apple leaves images dataset, which affirms the in-field applicability of the proposed framework. The superiority of the proposed model is further demonstrated through a comparative study with equivalent lightweight models.

Why it matches plant phenotyping methods植物の葉画像から病害状態を推定する軽量モバイル深層学習手法の開発・比較検証が中心であり、植物表現型（病徴・病害状態）の取得手法に該当する。

abstractwe propose a mobile-based lightweight deep learning-based model

The shape of aroma: measuring and modeling citrus oil gland distribution →

From preventing scurvy to being part of religious rituals, citrus are intrinsically connected to human health and perception. From tiny mandarins to head-sized pummelos, citrus capability of hybridization provides a vastly diverse array of fruit sizes and shapes, which in turn corresponds to a diversity of flavors and aromas. These sensory qualities are tightly linked to oil glands in the citrus skin. The oil glands are also key to understanding fruit development, and the essential oils contained by them are fundamental in the food and perfume industries. We study the shape of citrus based on 3D X-ray CT scan reconstruction of 163 different citrus samples comprising 58 different species and cultivars, including samples of all fundamental citrus species. First, using the power of X-rays and image processing, we are able to compare and contrast size ratios between different tissues, such as the size of the skin compared to the rind or the flesh. Second, we model the fruit shape as an ellipsoidal surface, and later we study and infer possible oil gland distributions on this surface using principles of directional statistics. We finally compare and contrast these overall fruit shape models along their gland distributions across different citrus species. This morphological modeling will allow us later to link genotype with phenotype, furthering our insight on how the physical shape is genetically specified in DNA.

Why it matches plant phenotyping methods3D X線CT、画像処理、形状モデリング、方向統計を中核として、柑橘果実の形態と油腺分布という植物形質を定量化しているため、フェノタイピング手法研究に該当します。

abstractWe study the shape of citrus based on 3D X-ray CT scan reconstruction of 163 different citrus samples comprising 58 different species and cultivars

Citrus fruits maturity detection in natural environments based on convolutional neural networks and visual saliency map →

Citrus fruits do not ripen at the same time in natural environments and exhibit different maturity stages on trees, hence it is necessary to realize selective harvesting of citrus picking robots. The visual attention mechanism reveals a physiological phenomenon that human eyes usually focus on a region that is salient from its surround. The degree to which a region contrasts with its surround is called visual saliency. This study proposes a novel citrus fruit maturity method combining visual saliency and convolutional neural networks to identify three maturity levels of citrus fruits. The proposed method is divided into two stages: the detection of citrus fruits on trees and the detection of fruit maturity. In stage one, the object detection network YOLOv5 was used to identify the citrus fruits in the image. In stage two, a visual saliency detection algorithm was improved and generated saliency maps of the fruits; The information of RGB images and the saliency maps were combined to determine the fruit maturity class using 4-channel ResNet34 network. The comparison experiments were conducted around the proposed method and the common RGB-based machine learning and deep learning methods. The experimental results show that the proposed method yields an accuracy of 95.07%, which is higher than the best RGB-based CNN model, VGG16, and the best machine learning model, KNN, about 3.14% and 18.24%, respectively. The results prove the validity of the proposed fruit maturity detection method and that this work can provide technical support for intelligent visual detection of selective harvesting robots.

Why it matches plant phenotyping methods画像と視覚サリエンシー、CNNを組み合わせて樹上果実の成熟度を推定する手法が研究の中心であり、収穫対象の単なる検出を超えて植物器官の状態を定量化している。

abstractThis study proposes a novel citrus fruit maturity method combining visual saliency and convolutional neural networks to identify three maturity levels of citrus fruits.

Plants Disease Image Classification Based on Lightweight Convolution Neural Networks →

Plants diseases is a major threat to agricultural production. Reduced yield due to plant diseases can lead to immeasurable economic losses. Therefore, the detection and classification of plant diseases are of great significance. Most of the existing plant disease detection methods focus on improving the identification accuracy. However, besides accuracy, real-time performance cannot be ignored. In this paper, a new module named 2-way residual dense layer is presented to effectively decrease the number of parameters in our network. In this module, depth separable convolution is introduced, which reduces the amount of parameter calculation and achieves a performance of over 98%. Our network is verified by an open dataset which includes 4503 images from four classes, including Mango, Arjun, Alstonia Scholaris, Guava, Bael, Jamun, Jatropha, Pongamia Pinnata, Basil, Pomegranate, Lemon, and Chinar. The leaf images of these plants have healthy and diseased condition. The experimental results showed that this method can be practically applied to the identification of plant leaf diseases and provide a basis for the identification of other leaf diseases.

Why it matches plant phenotyping methods植物葉画像から病害状態を分類する軽量CNN手法の開発・検証が研究の中心であり、植物の病害状態を直接推定するため、植物フェノタイピング手法として含める。

abstractIn this paper, a new module named 2-way residual dense layer is presented to effectively decrease the number of parameters in our network.

Towards Sensory Assessment Classification using Short-Wave NIR Spectroscopy for Orange Cultivars →

Abstract The global orange industry constantly faces new technical challenges to meet consumer demands for quality fruits. Instead of traditional subjective fruit quality assessment methods, the interest in the horticulture industry has increased in an objective, quantitative, and non-destructive assessment methods. Oranges have a thick peel which makes their non-destructive quality assessment challenging. This paper evaluates the potential of short-wave NIR spectroscopy and direct sweetness classification for Pakistani cultivars of orange i.e., Blood red, Mosambi, and Succari. The correlation between quality indices i.e., Brix, titratable acidity (TA), Brix: TA and BrimA (Brix minus acids), sensory assessment of the fruit, and short-wave NIR spectra is analyzed. Mix cultivar oranges are then classified as sweet, mixed, and acidic based on short-wave NIR spectra. Short-wave NIR spectral data were obtained using the industry standard F-750 fruit quality meter (310–1100 nm). Reference Brix and TA measurements were taken using standard destructive testing methods. Reference taste labels i.e., sweet, mix, and acidic, were acquired by sensory evaluation of samples. For indirect fruit classification, partial least squares regression models were developed for Brix, TA, Brix: TA, and BrimA estimation with a correlation coefficient of 0.57, 0.73, 0.66, and 0.55 respectively, on independent test data. For direct fruit classification, the ensemble classifier achieved 81.03% accuracy for 3 class (sweet, mix, and acidic) classification on independent test data. We observed a good correlation between NIR spectra and sensory assessment instead of quality indices. Hence, direct classification is more suitable for orange sweetness classification using NIR spectroscopy than the estimation of quality indices.

Why it matches plant phenotyping methodsオレンジ果実の甘味・品質という植物器官の形質を、短波長NIR分光で非破壊推定・分類する手法を開発および評価しており、表現型取得が研究の中心である。

abstractthe interest in the horticulture industry has increased in an objective, quantitative, and non-destructive assessment methods

Identification of citrus diseases based on AMSR and MF-RANet. →

Background As one of the most widely planted fruit trees in southern China, citrus occupies an important position in the agriculture field and forestry economy in China. There are many kinds of citrus diseases. If citrus infected with diseases cannot be controlled in time, it easily seriously affects citrus production and causes large economic losses. Timely monitoring of disease characteristics in the citrus growth process is important for implementing timely control measures. Citrus images are easily disturbed by environmental factors such as dust, low light, clouds or leaf shadows. This makes it easy for some disease spot features in citrus pictures to be obscured. Occluded lesions cannot be effectively extracted and recognized. Second, similar characteristics of different diseases also make it difficult to distinguish the different types of diseases. However, the existing machine vision technology for identifying citrus diseases still has some difficulties in dealing with the above problems. Results This paper proposes a new citrus disease identification framework. First, a citrus image enhancement algorithm based on the MSR-AMSR algorithm is proposed, which can enhance the image and highlight the disease characteristic information. The AMSR algorithm can also greatly alleviate the interference of clouds and low light on image lesions, making the image features clearer. Second, an MF-RANet network is proposed to recognize citrus disease images. MF-RANet is composed of a main feature frame and a detail feature frame. The main feature frame uses the cross stacking structure of ResNet50 and RAM to extract the main features in the citrus image dataset. RAM is used to extract the attention weight in the feature layer, which enables RAM to give higher weight to disease features. The detailed feature frame path uses AugFPN to extract features from multiple scales and fuse the main feature frame path. AugFPN enables the network to retain more detailed features, so it can effectively distinguish similar features in different diseases. In addition, we use the ELU activation function not only to solve the problem of gradient explosion and gradient disappearance but also to effectively use the negative input of the network. Finally, we use the label smoothing regularization method to prevent overfitting the network in the classification process. Finally, the experimental results show that the highest detection accuracy of the network for Huanglong disease, Corynespora blight of citrus, fat spot macular disease, citrus scab, citrus canker and healthy citrus is 96.77%, 96.22%, 95.96%, 95.93%, 94.04% and 97.55%, respectively. Conclusions The citrus disease algorithm based on AMSR and MF-RANet can effectively perform the disease detection function. It has a high recognition rate for different kinds of citrus diseases. With the addition of AMSR preprocessing, RAM, AugFPN, ELU activation function and other structures, the MF-RANet network performance improves.

Why it matches plant phenotyping methods柑橘の病斑・病害状態を画像から抽出・認識する画像処理および深層学習フレームワークの開発が研究の中心であり、植物病害フェノタイピング手法に該当する。

abstractThis paper proposes a new citrus disease identification framework.

Citrus greening disease recognition algorithm based on classification network using TRL-GAN →

The monitoring and prevention and control of citrus yellow dragon disease is a significant measure to ensure citrus production. If yellow dragon disease appears in citrus orchards, it will cause root rot, fruit deformation and wilting of fruit trees, which will eventually spread to every fruit tree in the whole orchard and cause the death of fruit trees, so it is very meaningful to detect the symptoms of citrus yellow dragon disease early and take appropriate treatment and prevention measures. Pratically, the orchard owner will remove the corresponding fruit trees as soon as they are found to be infected with Huanglong disease, so that it is extremely problematic to obtain a large number of Huanglong disease leaf data. Meanwhile, due to the uncertainty of the pathological trait distribution of citrus yellow dragon disease leaves and the extreme shortage of data, the convolutional neural network model learned in a small number of samples is not capable of generalization. In order to improve the accuracy and generalization of Citrus Greening Disease recognition algorithm, this paper introduces Texture Reconstruction Loss CycleGAN(TRL-GAN) to generate citrus diseased leaf data in realistic scene to increase the richness of samples, and thus proposes the Recognizing Citrus Greening Based on TRL-GAN(RCG TRL-GAN). This algorithm firstly performs background culling by using the instance segmentation network Mask RCNN for realistic scenes citrus yellow dragon disease mottled, zinc deficiency, magnesium deficiency, leaf veins yellowing and other corresponding symptomatic leaves, then introduces texture reconstruction loss improvement CycleGAN as training and migrates the diseased leaf style to ordinary green leaves for data expansion, and finally uses the expanded dataset to train the convolutional neural network. Experimental results on the constructed dataset of 4516 images (762 mottled, 749 Zn deficient, 737 Mg deficient, 721 Vein yellowing, 783 Diachyma yellowing, 764 green leaves) reveal that TRL-GAN has 13.49% and 1.1% improvement in FID and KID, respectively, relative to the original structure CycleGAN, and has been identified by six citrus yellow dragon disease experts and three vision professionals identify that the fake data generated by TRL-GAN have similarity with the leaf pathological characteristics and real data, and also by using T-SNE technique it is observed that the real data have similar distribution with the generated fake data in two-dimensional plane. Meanwhile, the more outstanding accuracy performance in the classification network is ResNeXt101 with 97.45% accuracy, and the average accuracy of RCG TRL-GAN technique in the recognition of classification network is improved 2.76%. The study proves that the RCG TRL-GAN effectively improves the citrus greening disease phenotype data generation and recognition, and can provide method reference for the expansion and recognition of complex plant disease phenotype images.

Why it matches plant phenotyping methods柑橘病害症状画像を対象に、セグメンテーション、GANによる病徴画像生成、分類認識を統合した植物病害フェノタイピング手法を開発・評価しており、方法自体が中心である。

abstractthis paper introduces Texture Reconstruction Loss CycleGAN(TRL-GAN) to generate citrus diseased leaf data in realistic scene to increase the richness of samples

Citrus Huanglongbing detection and semi-quantification of the carbohydrate concentration based on micro-FTIR spectroscopy. →

Citrus Huanglongbing (HLB) is nowadays one of the most fatal citrus diseases worldwide. Once the citrus tree is infected by the HLB disease, the biochemistry of the phloem region in midribs would change. In order to investigate the carbohydrate changes in phloem region of citrus midrib, the semi-quantification models were established to predict the carbohydrate concentration in it based on Fourier transform infrared microscopy (micro-FTIR) spectroscopy coupled with chemometrics. Healthy, asymptomatic-HLB, symptomatic-HLB, and nutrient-deficient citrus midribs were collected in this study. The results showed that the intensity of the characteristic peak varied with the carbohydrate (starch and soluble sugar) concentration in citrus midrib, especially at the fingerprint regions of 1175-900 cm -1 , 1500-1175 cm -1 , and 1800-1500 cm -1 . Furthermore, semi-quantitative prediction models of starch and soluble sugar were established using the full micro-FTIR spectra and selected characteristic wavebands. The least squares support vector machine regression (LS-SVR) model combined with the random frog (RF) algorithm achieved the best prediction result with the determination coefficient of prediction ([Formula: see text]) of 0.85, the root mean square error of prediction (RMSEP) of 0.36%, residual predictive deviation (RPD) of 2.54, and [Formula: see text] of 0.87, RMSEP of 0.37%, RPD of 2.76, for starch and soluble sugar concentration prediction, respectively. In addition, multi-layer perceptron (MLP) classification models were established to identify HLB disease, achieving the overall classification accuracy of 94% and 87%, based on the full-range spectra and the optimal wavenumbers selected by the random frog (RF) algorithm, respectively. The results demonstrated that micro-FTIR spectroscopy can be a valuable tool for the prediction of carbohydrate concentration in citrus midribs and the detection of HLB disease, which would provide useful guidelines to detect citrus HLB disease.

Why it matches plant phenotyping methodsmicro-FTIRスペクトロスコピーと機械学習モデルを用いて、柑橘葉脈の糖濃度という生理形質とHLB病害状態を推定・分類する手法を開発し、予測性能と分類精度を評価しているため、方法が研究の中心である。

abstractthe semi-quantification models were established to predict the carbohydrate concentration in it based on Fourier transform infrared microscopy (micro-FTIR) spectroscopy coupled with chemometrics.

Two-wavelength image detection of early decayed oranges by coupling spectral classification with image processing →

Citrus decay is one of the most serious postharvest diseases, which can cause great economic losses and food safety problem. Early decay has no obvious features, which makes rapid detection and grading of decayed citrus fruit a major challenge for citrus industry. This study utilized the unique image-spectrum fusion property of hyperspectral imaging, and proposed a strategy to quickly identify the citrus with early decay using only two wavelength images. The typical average spectra of sound and decayed tissues were extracted. The linear partial least squares-discriminant analysis (PLS-DA) and nonlinear back-propagation artificial neural network (BP-ANN) models were constructed for classifying two types of tissues. MC-UVE-SPA algorithm by combining Monte Carlo cross-validation (MC-UVE) with successive projections algorithm (SPA) was used to extract 16 variables characterizing two types of tissues. The wavelength images corresponding to the extracted variables were performed principal component analysis (PCA) to find the optimal PC image. Only two wavelength images at 568.8 nm and 771.2 nm were selected by analyzing the weighting coefficients of the third principal component (PC3) image. An improved watershed segmentation method was proposed to segment decay region in oranges based on PC2 image of the selected two wavelength images. Classification performance of the proposed algorithm was evaluated by all samples. The results showed that the overall classification accuracy of 96.6% was achieved, with 100% and 91.3% for the decayed and sound oranges respectively. The proposed detection strategy only involves two wavelength images, which will contribute to the establishment of a fast and low-cost multispectral imaging system for detection of citrus with early decay.

Why it matches plant phenotyping methodsオレンジ果実の腐敗状態を二波長画像、スペクトル分類、画像処理で抽出する検出法を開発し、分類精度を評価しており、植物器官の病害状態取得が中心である。

abstractproposed a strategy to quickly identify the citrus with early decay using only two wavelength images

DS-MENet for the classification of citrus disease. →

Affected by various environmental factors, citrus will frequently suffer from diseases during the growth process, which has brought huge obstacles to the development of agriculture. This paper proposes a new method for identifying and classifying citrus diseases. Firstly, this paper designs an image enhancement method based on the MSRCR algorithm and homomorphic filtering algorithm optimized by Laplacian (HFLF-MS) to highlight the disease characteristics of citrus. Secondly, we designed a new neural network DS-MENet based on the DenseNet-121 backbone structure. In DS-MENet, the regular convolution in Dense Block is replaced with depthwise separable convolution, which reduces the network parameters. The ReMish activation function is used to alleviate the neuron death problem caused by the ReLU function and improve the robustness of the model. To further enhance the attention to citrus disease information and the ability to extract feature information, a multi-channel fusion backbone enhancement method (MCF) was designed in this work to process Dense Block. We use the 10-fold cross-validation method to conduct experiments. The average classification accuracy of DS-MENet on the dataset after adding noise can reach 95.02%. This shows that the method has good performance and has certain feasibility for the classification of citrus diseases in real life.

Why it matches plant phenotyping methods柑橘病害の画像特徴抽出と分類モデルを中心に開発・検証しており、植物の病害状態を直接推定するフェノタイピング手法である。

abstractThis paper proposes a new method for identifying and classifying citrus diseases.

Soluble Solids Content Binary Classification of Miyagawa Satsuma in Chongming Island Based on Near Infrared Spectroscopy. →

Citrus is one of the most important fruits in China. Miyagawa Satsuma, one kind of citrus, is a nutritious agricultural product with regional characteristics of Chongming Island. Near-infrared Spectroscopy (NIR) is a proper method for studying the quality of fruits, because it is low-cost, efficient, non-destructive, and repeatable. Therefore, the NIR technique is used to detect citrus's soluble solid content (SSC) in this study. After obtaining the original spectral data, the first 70% of them are divided into the training set and 30% into the test set. Then, the Random Frog algorithm is chosen to select characteristic wavelengths, which reduces the dimension of the data and the complexity of the model, and accordingly makes the generalization of the classification model better. After comparing the performance of various classifiers (AdaBoost, KNN, LS-SVM, and Bayes) under different characteristic wavelength numbers, the AdaBoost classifier outperforms using 275 characteristic wavelengths for modeling eventually. The accuracy, precision, recall, and F 1 -score are 78.3%, 80.5%, 78.3%, and 0.780, respectively and the ROC (Receiver Operating Characteristic Curve, ROC curve) is close to the upper left corner, suggesting that the classification model is acceptable. The results demonstrate that it is feasible to use the NIR technique to estimate whether the citrus is sweet or not. Furthermore, it is beneficial for us to apply the obtained models for identifying the quality of citrus correctly. For fruit traders, the model helps them to determine the growth cycle of citrus more scientifically, improve the level of citrus cultivation and management and the final fruit quality, and thus increase the economic income of fruit traders.

Why it matches plant phenotyping methodsNIR分光で柑橘果実の可溶性固形分（糖度）を非破壊推定し、波長選択と複数分類器を比較・評価しており、植物形質取得法が中心である。

abstractTherefore, the NIR technique is used to detect citrus's soluble solid content (SSC) in this study.

Estimating the distribution of chlorophyll content in CYVCV infected lemon leaf using hyperspectral imaging →

Yellow vein clearing disease of lemon caused by the citrus yellow vein clearing virus (CYVCV) is a new devastating disease in lemon. Chlorophyll contents and distributions are critical biochemical parameters to assess this new disease. The objective of this study was to develop an efficient method for detecting chlorophyll contents and distributions in lemon leaves infected with CYVCV. Spectral reflectance and chlorophyll content data of healthy, nitrogen deficient, pesticide-damaged and the corresponding CYVCV-infected leaves were obtained. Successive projections algorithm, random frog, competitive adaptive reweighted sampling (CARS), synergy interval partial least squares and synergy interval partial least squares combined with successive projections algorithm were applied for reducing data dimension, respectively. Prediction models were built by applying least squares-support vector machine algorithm (LS-SVM). The results showed that the characteristic wavelengths were basically located in “green peak” (near 550 nm), “red valley” (near 680 nm), and “red edge” (680–750 nm). Prediction model based on the wavelengths extracted by CARS achieved the optimal prediction results with determination coefficient of 0.94, relative prediction deviation of 3.91 and root mean square error of 0.10 in testing set. Thus, the CARS algorithm combined with LS-SVM was applied for distributions of chlorophyll contents. The chlorophyll content of each pixel of the six types of leaves were displayed, and significantly reduced in the mesophyll tissue near the veins of CYVCV-infected leaves. These results suggest that the hyperspectral imaging coupled with CARS and LS-SVM can be used to efficiently measure the distribution of chlorophyll content in CYVCV infected lemon leaf, which provides a reference for a better understanding of the symptoms of disease.

Why it matches plant phenotyping methods感染葉のクロロフィル含量・分布を hyperspectral imaging と CARS/LS-SVM で推定する手法の開発が中心であり、植物の病害状態に関連する生理形質を直接測定している。

abstractThe objective of this study was to develop an efficient method for detecting chlorophyll contents and distributions in lemon leaves infected with CYVCV.

Noise-tolerant RGB-D feature fusion network for outdoor fruit detection →

In the process of farm automation, fruit detection is the basis and guarantee for yield prediction, automatic picking, and other orchard operations. RGB images can only obtain the two-dimensional information of the scene, which is not sufficient to effectively distinguish fruits that are dense growth and occlusion by branches and leaves. With the development of depth sensors, using RGB-D images with more complementary information can boost the performance of fruit detection. However, due to the nature of sensors and scene configurations, the quality of outdoor depth images is poor, posing a challenge when fusing RGB-D features. Therefore, this paper proposes an end-to-end RGB-D object detection network, termed as noise-tolerant feature fusion network (NT-FFN), to utilize the outdoor multi-modal data properly and improve the detection accuracy. Specifically, the NT-FFN first uses two structurally identical feature extractors to extract single-modal (color and depth) features, which is the base of the subsequent feature fusion. Then, to avoid introducing too much depth noise and focus the perception on the important part of the features, an attention-based fusion module is designed to adaptively fuse the multi-modal features. Finally, multi-scale features from the color images and the fusion modules are used to predict object position, which not only improves the network's ability to detect multi-scale objects but also further enhances the noise immunity of the network. In addition, this paper constructs an RGB-D citrus fruit dataset, which contributes to comprehensively evaluating the proposed network. Evaluation metrics on the dataset show that the NT-FFN achieves an AP⁵⁰ of 95.4% with a real-time speed, which outperforms single-modal methods, common multi-modal fusion strategies, and advanced multi-modal detection methods. The proposed NT-FFN also achieves excellent detection results in other fruit detection tasks, which verifies its generalization ability. This study provides the possibility and foundation for performing multi-modal information fusion in outdoor fruit detection.

Why it matches plant phenotyping methodsRGB-D画像から果実の位置を検出する手法を開発し、独自データセットで評価しているため、植物器官の観測・抽出方法が中心的である。

abstractTherefore, this paper proposes an end-to-end RGB-D object detection network, termed as noise-tolerant feature fusion network (NT-FFN), to utilize the outdoor multi-modal data properly and improve the detection accuracy.

Detection of citrus Huanglongbing (HLB) based on the HLB-induced leaf starch accumulation using a home-made computer vision system →

Citrus Huanglongbing (HLB) is a highly destructive and transmissible citrus disease spread by insects. As it has no effective treatment, there is an urgent need for a rapid HLB detection method that can be applied in orchards to remove the diseased trees. In this research, the contribution of two symptoms, based on HLB-induced starch accumulation, were evaluated to HLB detection using a home-made computer vision system with two imaging modes. The reflection imaging mode was used to detect blotchy mottles symptoms. A 660 nm light source was used and the absorption of different wavelengths by chlorophyll and lutein was measured. The transmission imaging was designed to detect abnormal starch accumulation within the leaf. An image was taken with 590 nm polarised light penetrating the leaf, then the angle of linear polarisation (AoLP) was calculated by the Stokes vector. The AoLP reflects the ability of the internal leaf components to rotate polarised light. The polarisation angle reflects the starch content of the leaves and determines their disease status. Multi-layer perceptron (MLP), random forest (RF), and logistic regression (LR) classifiers were then evaluated. The RF classifier performed better in the reflection experiment, with a classification accuracy of 96.67%. In the transmission experiment, the LR classifier had 88.33% recognition rate. The results show that HLB-induced starch accumulation contributes to the visual detection of HLB. This research supports the high accuracy and portability of HLB detection equipment and it has great practical importance for preventing and controlling of HLB disease.

Why it matches plant phenotyping methods柑橘葉の症状・デンプン蓄積を画像および偏光計測で取得し、HLB病状態を分類するコンピュータビジョン手法と装置を開発・評価しており、植物表現型取得が研究の中心である。

abstracta rapid HLB detection method that can be applied in orchards

Real-Time Assessment of Mandarin Crop Water Stress Index. →

The use of plant-based indicators and other conventional means to detect the level of water stress in crops may be challenging, due to their difficulties in automation, their arduousness, and their time-consuming nature. Non-contact and non-destructive sensing methods can be used to detect the level of water stress in plants continuously and to provide automatic sensing and controls. This research aimed at determining the viability, efficiency, and swiftness in employing the commercial Workswell WIRIS Agro R infrared camera (WWARIC) in monitoring water stress and scheduling appropriate irrigation regimes in mandarin plants. The experiment used a four-by-three randomized complete block design with 80−100% FC water treatment as full field capacity and three deficit irrigation treatments at 70−75% FC, 60−65% FC, and 50−55% FC. Air temperature, canopy temperature, and vapor pressure deficits were measured and employed to deduce the empirical crop water stress index, using the Idso approach (CWSI(Idso)) as well as baseline equations to calculate non-water stress and water stressed conditions. The relative leaf water content (RLWC) of mandarin plants was also determined for the growing season. From the experiment, CWSI(Idso) and CWSI were estimated using the Workswell Wiris Agro R infrared camera (CWSIW) and showed a high correlation (R2 = 0.75 at p < 0.05) in assessing the extent of water stress in mandarin plants. The results also showed that at an altitude of 12 m above the mandarin canopy, the WWARIC was able to identify water stress using three modes (empirical, differential, and theoretical). The WWARIC’s color map feature, presented in real time, makes the camera a suitable device, as there is no need for complex computations or expert advice before determining the extent of the stress the crops are subjected to. The results prove that this novel use of the WWARIC demonstrated sufficient precision, swiftness, and intelligibility in the real-time detection of the mandarin water stress index and, accordingly, assisted in scheduling irrigation.

Why it matches plant phenotyping methods赤外線カメラによる作物水ストレス指数の非接触・リアルタイム推定が研究の中心であり、従来のCWSIとの相関検証と実用性評価を行っている。

abstractThis research aimed at determining the viability, efficiency, and swiftness in employing the commercial Workswell WIRIS Agro R infrared camera (WWARIC) in monitoring water stress and scheduling appropriate irrigation regimes in mandarin plants.

Spectrum classification of citrus tissues infected by fungi and multispectral image identification of early rotten oranges. →

Citrus fruit is susceptible to postharvest rot by fungal infection. The detection of early rot is difficult due to similar skin characteristics to sound area, which limits the ability of the grading system to evaluate the comprehensive quality of citrus. In this study, the visible and near infrared hyperspectral imaging system with the wavelength range of 325-1000 nm was used to collect hyperspectral images of oranges. Hyperspectral data of three types of tissues including sound tissue from 80 samples, rotten tissue infected by Penicillium digitatum from 100 samples and rotten tissue infected by Penicillium italicum from 100 samples were extracted. The bootstrapping soft shrinkage (BOSS) and BOSS-SPA (BOSS-Successive Projections Algorithm) combination algorithm were separately used to optimize spectrum variables. The partial least squares discriminant analysis (PLS-DA) model for classifying three types of tissues and PLS-DA model for classifying two types of tissues (sound tissue and rotten tissue) were constructed based on full-spectrum and the selected informative variables. Model comparisonshowed that the BOSS-PLS-DA model can effectively identify three types of tissues with the classification accuracy of 97.1%, while the BOSS-SPA-PLS-DA model was more effective for the binary classification of sound and rotten citrus tissues with the accuracy of 100%. Furthermore, the wavelength images corresponding to the nine informative variables extracted by BOSS-SPA were performed the principal component analysis (PCA), and four feature wavelength images (508, 568, 578 and 614 nm) were obtained by analyzing the weighting coefficients of each single-wavelength images constituting the optimal principal component (PC) image. Finally, a fast multispectral image processing algorithm combined with the global threshold theory was proposed for the rotten orange detection based on the extracted four wavelength images. A total of 280 samples including 80 sound and 200 rotten samples were used to evaluate the classification ability, which showed the proposed multispectral image detection algorithm can successfully differentiate between sound and rotten oranges with an overall classification accuracy of 98.6%.

Why it matches plant phenotyping methods柑橘果实腐敗組織という植物状態を対象に、ハイパースペクトル画像、特徴波長抽出、画像処理アルゴリズムを開発・評価しており、病害状態の表現型取得が研究の中心です。

abstractthe visible and near infrared hyperspectral imaging system with the wavelength range of 325-1000 nm was used to collect hyperspectral images of oranges.

Diagnosis of HLB-asymptomatic citrus fruits by element migration and transformation using laser-induced breakdown spectroscopy. →

Huanglongbing (HLB) is one of the most devastating bacterial diseases in citrus growth and there is no cure for it. The mastery of elemental migration and transformation patterns can effectively analyze the growth of crops. The law of element migration and transformation in citrus growth is not very clear. In order to obtain the law of element migration and transformation, healthy and HLB-asymptomatic navel oranges collected in the field were taken as research objects. Laser-induced breakdown spectroscopy (LIBS) is an atomic spectrometry technique for material component analysis. By analyzing the element composition of fruit flesh, peel and soil, it can know the specific process of nutrient exchange and energy exchange between plants and the external environment, as well as the rules of internal nutrient transportation, distribution and energy transformation. Through the study of elemental absorption, the growth of navel orange can be effectively monitored in real time. HLB has an inhibitory effect on the absorption of navel orange. In order to improve the detection efficiency, LIBS coupled with SVM algorithms was used to distinguish healthy navel oranges and HLB-asymptomatic navel oranges. The classification accuracy was 100%. Compared with the traditional detection method, the detection efficiency of LIBS technology is significantly better than the polymerase chain reaction method, which provides a new means for the diagnosis of HLB-asymptomatic citrus fruits.

Why it matches plant phenotyping methodsLIBSとSVMを用いて、柑橘果実の元素情報から無症状HLB感染という植物の病態を判別し、PCRとの比較および精度評価を行っている。病害状態の取得・判定法が研究の中心である。

abstractLIBS coupled with SVM algorithms was used to distinguish healthy navel oranges and HLB-asymptomatic navel oranges.

Accurate nondestructive prediction of soluble solids content in citrus by near‐infrared diffuse reflectance spectroscopy with characteristic variable selection →

This study employs a near‐infrared diffuse reflectance spectroscopy (NIRDRS) system to accurate nondestructive determine citrus soluble solids content (SSC). The penetration experiment results showed that the interference of thick peel is large and NIRDRS light has the ability to penetrate the peel to a certain extent. Partial least squares with different characteristic variable selection methods were used to establish the quantitative model of SSC. The results demonstrated that characteristic variable selection methods can select targeted characteristic variables and improve the accuracy with few variables. Monte Carlo‐uninformative variable elimination method was selected as the optimal prediction performance. In the best prediction model, the correlation coefficient and root mean square error of prediction of the prediction set are 0.854 and 0.7%Brix, respectively, while the variable number decreases to 440 from 1557. Furthermore, the models using the average spectra of four points on the equator are the most appropriate. NOVELTY IMPACT STATEMENT: The penetrating ability of near‐infrared diffuse reflectance spectroscopy (NIRDRS) light to thick peel and the prediction of internal quality of citrus is still unsatisfactory with traditional partial least squares (PLS) algorithm due to the interference of thick peel. In this study, a nondestructive method for the analysis of soluble solids content (SSC) in citrus was established by NIRDRS with characteristic variable selection algorithms. The results demonstrated that NIRDRS light has the ability to penetrate the peel to a certain extent, while characteristic variable selection methods can select targeted characteristic variables and improve the accuracy of quantitative analysis models with fewer variables.

Why it matches plant phenotyping methods柑橘果実の可溶性固形分をNIR分光で非破壊推定する手法を開発・検証しており、植物器官の品質形質取得が中心である。

abstractThis study employs a near‐infrared diffuse reflectance spectroscopy (NIRDRS) system to accurate nondestructive determine citrus soluble solids content (SSC).

Rapid Evaluation of the Resistance of Citrus Germplasms Against Xanthomonas citri subsp. citri . →

Xanthomonas citri subsp. citri ( Xcc ) is the causal agent of citrus bacterial canker (CBC), one of the most devastating citrus diseases. Most commercial citrus varieties are susceptible to CBC. However, some citrus varieties and wild citrus germplasms are CBC resistant and are promising in genetic increases in citrus resistance against CBC. We aimed to evaluate citrus germplasms for resistance against CBC. First, we developed a rapid evaluation method based on enhanced yellow fluorescent protein (eYFP)-labeled Xcc . The results demonstrated that eYFP does not affect the growth and virulence of Xcc . Xcc-eYFP allows measurement of bacterial titers but is more efficient and rapid than the plate colony counting method. Next, we evaluated citrus germplasms collected in China. Based on symptoms and bacterial titers, we identified that two citrus germplasms ('Ichang' papeda and 'Huapi' kumquat) are resistant, whereas eight citrus germplasms ('Chongyi' wild mandarin, 'Mangshan' wild mandarin, 'Ledong' kumquat, 'Dali' citron, 'Yiliang' citron, 'Longyan' kumquat, 'Bawang' kumquat, and 'Daoxian' wild mandarin) are tolerant. In summary, we have developed a rapid evaluation method to test the resistance of citrus plants against CBC. This method was successfully used to identify two highly canker-resistant citrus germplasms and eight citrus germplasms with canker tolerance. These results could be leveraged in traditional breeding contexts or be used to identify canker resistance genes to increase the disease resistance of commercial citrus varieties via biotechnological approaches.

Why it matches plant phenotyping methods植物のかんきつ病害抵抗性を迅速評価する方法を開発し、症状と細菌力価に基づく抵抗性・耐性評価へ適用しており、病害表現型の取得が中心です。

abstractIn summary, we have developed a rapid evaluation method to test the resistance of citrus plants against CBC.

Combining multicolor fluorescence imaging with multispectral reflectance imaging for rapid citrus Huanglongbing detection based on lightweight convolutional neural network using a handheld device →

Citrus Huanglongbing (HLB) has posed a great challenge to the citrus production. Timely removal of HLB infected trees was considered as one of the most effective strategies for citrus orchard management. Therefore, rapid detection of HLB disease is urgently needed. The study was aimed to propose an effective method for HLB disease detection by developing a handheld device to capture multicolor fluorescence and multispectral reflectance images synchronously. Additionally, the deep learning and transfer learning technologies were introduced for citrus HLB disease detection. The results demonstrated that the lightweight convolutional neural network (MobileNetV3) can achieve an overall accuracy of 92.1% with the false negative rate of 12.1% at epochs of 33 by combining multicolor fluorescence with multispectral reflectance images as the input of MobileNetV3 model using the dataset of Navel orange. It implied that structural and physiological information from reflectance images and multicolor fluorescence images relating to photosynthesis and secondary metabolites were valuable for rapid HLB disease detection. The transfer learning method of fine-tuning model obtained a superior transferring ability than that of reuse-model with the overall accuracy of 96.5% for Ponkan. These results demonstrated the feasibility of developed handheld device based on multicolor fluorescence and multispectral reflectance imaging combined with deep learning and transfer learning technologies for high throughput HLB disease detection in different infected statuses and cultivars.

Why it matches plant phenotyping methods柑橘のHLB感染状態を、携帯型の蛍光・マルチスペクトル画像と深層学習で植物症状・生理状態として推定する取得・解析手法を開発しており、方法が中心的である。

abstractThe study was aimed to propose an effective method for HLB disease detection by developing a handheld device to capture multicolor fluorescence and multispectral reflectance images synchronously.

Classification of Citrus Diseases Using Optimization Deep Learning Approach. →

Most plant diseases have apparent signs, and today's recognized method is for an expert plant pathologist to identify the disease by looking at infected plant leaves using a microscope. The fact is that manually diagnosing diseases is time consuming and that the effectiveness of the diagnosis is related to the pathologist's talents, making this a great application area for computer-aided diagnostic systems. The proposed work describes an approach for detecting and classifying diseases in citrus plants using deep learning and image processing. The main cause of decreased productivity is considered to be plant diseases, which results in financial losses. Citrus is an important source of nutrients such as vitamin C all around the world. On the contrary, citrus diseases have a negative impact on the citrus fruit and quality. In the recent decade, computer vision and image processing techniques have become increasingly popular for the detection and classification of plant diseases. The suggested approach is evaluated on the citrus disease image gallery dataset and the combined dataset (citrus image datasets of infested scale and plant village). These datasets were used to identify and classify citrus diseases such as anthracnose, black spot, canker, scab, greening, and melanose. AlexNet and VGG19 are two kinds of convolutional neural networks that were used to build and test the proposed approach. The system's total performance reached 94% at its best. The proposed approach outperforms the existing methods.

Why it matches plant phenotyping methods柑橘葉の画像から病害状態を検出・分類するコンピュータビジョン手法が研究の中心であり、植物病害フェノタイピングに該当する。

abstractThe proposed work describes an approach for detecting and classifying diseases in citrus plants using deep learning and image processing.

Assessment of state-of-the-art deep learning based citrus disease detection techniques using annotated optical leaf images →

Citrus (Citrus reticulata) plants are affected by several diseases and require keen attention to detect and cure the diseases in time; otherwise, significant financial loss is incurred. With the advancement of computer vision and deep learning techniques, identifying various diseases is becoming simpler. However, this process requires a proper dataset of infected leaves and a suitable detector to recognise the diseases. Because the publicly available citrus leaf datasets are not annotated, they are not suited for disease detection tasks. Therefore, a new dataset (called CCL’20) comprising images of infected citrus leaves with multiple classes of diseases, including precise annotations, is developed. Primarily, machine learning models are used in plant disease detection, and only limited deep learning models are utilised in agricultural applications. This paper has identified the CNN based detectors best suited for agricultural engineering, such as CenterNet, YOLOv4, Faster-RCNN, DetectoRS, Cascade-RCNN, Foveabox and Deformabe Detr, implemented and fine-tuned them to detect citrus leaf diseases using our CCL’20 dataset. Extensive performance and computational analysis is carried out to determine how effectively these models diagnose different stages of citrus leaf diseases. This paper presents the state-of-the-art CNN detectors for citrus leaf disease detection, evaluated based on their precision, recall, and other valuable parameters such as training parameters, inference time, memory usage, speed and accuracy trade-off for each model. The results show that the Scaled YOLOv4 P7 achieves fast and early prediction of the diseases, and CenterNet2 with Res2Net 101 DCN-BiFPN predicts the early stage of citrus leaf diseases with high accuracy to other recent and efficient detecting models.

Why it matches plant phenotyping methods感染葉の画像から植物病害の状態を検出するデータセットを構築し、複数のCNN検出器を実装・比較・性能評価しており、植物表現型取得・抽出法が中心である。

abstractTherefore, a new dataset (called CCL’20) comprising images of infected citrus leaves with multiple classes of diseases, including precise annotations, is developed.

Spectral pattern study of citrus black rot caused by Alternaria alternata and selecting optimal wavelengths for decay detection. →

Fungal decay is one of the most common diseases that affect postharvest operations and sales of citrus. Sometimes, fungal disease develops and spreads inside the fruit and in the advanced stages of the disease, it appears apparent, so the use of efficient and reliable methods for early detection of the disease is very important. In this study, early detection of citrus black rot disease caused by Alternaria genus fungus was examined using spectroscopy. Jaffa oranges were inoculated with Alternaria alternata . The samples were inspected by spectroscopy (200-1100 nm) in the 1st, 2nd, and 3rd weeks after inoculation. The classification of healthy and infected samples and selection of most important wavelengths were conducted by soft independent modeling of class analogy (SIMCA). The most important wavelengths in the detection of healthy and infected samples of the 1st week were 507, 933, 937, and 950 nm with a classification accuracy of 60%. The most important wavelengths of the 2nd week were 522 and 787 nm with a classification accuracy of 60%. Also, wavelengths of 546, 660, 691, and 839 were found to be effective in the 3rd week with a classification accuracy of 100%.

Why it matches plant phenotyping methods柑橘果実の感染状態を分光法で検出・分類し、重要波長と分類精度を評価しており、植物病害状態の取得手法が研究の中心である。

abstractearly detection of citrus black rot disease caused by Alternaria genus fungus was examined using spectroscopy.

Green Citrus Detection and Counting in Orchards Based on YOLOv5-CS and AI Edge System. →

Green citrus detection in citrus orchards provides reliable support for production management chains, such as fruit thinning, sunburn prevention and yield estimation. In this paper, we proposed a lightweight object detection YOLOv5-CS (Citrus Sort) model to realize object detection and the accurate counting of green citrus in the natural environment. First, we employ image rotation codes to improve the generalization ability of the model. Second, in the backbone, a convolutional layer is replaced by a convolutional block attention module, and a detection layer is embedded to improve the detection accuracy of the little citrus. Third, both the loss function CIoU (Complete Intersection over Union) and cosine annealing algorithm are used to get the better training effect of the model. Finally, our model is migrated and deployed to the AI (Artificial Intelligence) edge system. Furthermore, we apply the scene segmentation method using the "virtual region" to achieve accurate counting of the green citrus, thereby forming an embedded system of green citrus counting by edge computing. The results show that the mAP@.5 of the YOLOv5-CS model for green citrus was 98.23%, and the recall is 97.66%. The inference speed of YOLOv5-CS detecting a picture on the server is 0.017 s, and the inference speed on Nvidia Jetson Xavier NX is 0.037 s. The detection and counting frame rate of the AI edge system-side counting system is 28 FPS, which meets the counting requirements of green citrus.

Why it matches plant phenotyping methods緑色の柑橘果実数を画像から検出・計数するYOLOv5-CSモデルとエッジシステムを開発しており、果実数という植物器官形質の取得方法が研究の中心である。

abstractwe proposed a lightweight object detection YOLOv5-CS (Citrus Sort) model to realize object detection and the accurate counting of green citrus in the natural environment.

Indirect quantitative analysis of soluble solid content in citrus by the leaves using hyperspectral imaging combined with machine learning. →

Due to the effect of bagging on fruit growth, non-destructive and in situ soluble solid content (SSC) in citrus detection remains a challenge. In this work, a new method for accurately quantifying SSC in citrus using hyperspectral imaging of citrus leaves was proposed. Sixty-five Ehime Kashi No. 28 citruses with surrounding leaves picked at two different times were picked for the experiment. Using the principal components analysis combined with Gaussian process regression model, the correlation coefficients of prediction-real value by citrus and its leaves in cross-validation were 0.972 and 0.986, respectively. In addition, the relationship between citrus leaves and SSC content was further explored, and the possible relationship between chlorophyll in leaves and SSC of citrus was analyzed. Comparing the quantitative analysis results by citrus and its leaves, the results show that the proposed method is a non-destructive and reliable method for determining the SSC by citrus leaves and has broad application prospects in indirect detection of citrus.

Why it matches plant phenotyping methods柑橘葉のハイパースペクトル画像と機械学習により果実SSCを非破壊推定する手法の提案・評価が中心であり、植物形質の取得手法に該当する。

abstracta new method for accurately quantifying SSC in citrus using hyperspectral imaging of citrus leaves was proposed

Citrus Huanglongbing Detection Based on Multi-Modal Feature Fusion Learning. →

Citrus Huanglongbing (HLB), also named citrus greening disease, occurs worldwide and is known as a citrus cancer without an effective treatment. The symptoms of HLB are similar to those of nutritional deficiency or other disease. The methods based on single-source information, such as RGB images or hyperspectral data, are not able to achieve great detection performance. In this study, a multi-modal feature fusion network, combining a RGB image network and hyperspectral band extraction network, was proposed to recognize HLB from four categories (HLB, suspected HLB, Zn-deficient, and healthy). Three contributions including a dimension-reduction scheme for hyperspectral data based on a soft attention mechanism, a feature fusion proposal based on a bilinear fusion method, and auxiliary classifiers to extract more useful information are introduced in this manuscript. The multi-modal feature fusion network can effectively classify the above four types of citrus leaves and is better than single-modal classifiers. In experiments, the highest accuracy of multi-modal network recognition was 97.89% when the amount of data was not very abundant (1,325 images of the four aforementioned types and 1,325 pieces of hyperspectral data), while the single-modal network with RGB images only achieved 87.98% recognition and the single-modal network using hyperspectral information only 89%. Results show that the proposed multi-modal network implementing the concept of multi-source information fusion provides a better way to detect citrus HLB and citrus deficiency.

Why it matches plant phenotyping methodsRGB画像とハイパースペクトル情報から柑橘葉の病害・栄養欠乏状態を推定するマルチモーダル手法を提案・評価しており、植物状態の取得と分類が研究の中心である。

abstracta multi-modal feature fusion network, combining a RGB image network and hyperspectral band extraction network, was proposed to recognize HLB from four categories (HLB, suspected HLB, Zn-deficient, and healthy).

Early detection of Huanglongbing with EESI-MS indicates a role of phenylpropanoid pathway in citrus. →

Huanglongbing (HLB), a devastating disease for citrus worldwide, is caused by Candidatus Liberibacter asiaticus (CLas). In this study, we employed a novel extractive electrospray ionization-mass spectrometry (EESI-MS) method to analyze the metabolites in leaves of uninfected and HLB-infected Newhall navel orange. The results showed that uninfected and HLB-infected leaves could be readily distinguished based on EESI-MS combined by multivariable analysis. Nine phenolic compounds involved in phenylpropanoid pathway, such as p-coumaric acid, naringin, and apigenin, were principal components to distinguish the leaves of uninfected and HLB-infected Newhall navel orange. Gene expression was also conducted to further explore the molecular mechanism of phenylpropanoid branch pathway in HLB. The expression of genes (4CL, HCT, CHI, CHS, CYP, and C12R) involved in phenylpropanoid branch pathway was increased in asymptomatic and early period of HLB-infected leaves, while decreased in later period of HLB-infected leaves. This study provides a novel method for early detection of citrus HLB and suggests the regulation mechanism of phenylpropanoid pathway in the interaction between citrus and CLas.

Why it matches plant phenotyping methodsEESI-MSを用いて感染葉と非感染葉を識別し、柑橘のHLB感染状態を早期検出する分析法を開発・適用しており、植物の病態判定が中心的な方法論的貢献である。

abstractwe employed a novel extractive electrospray ionization-mass spectrometry (EESI-MS) method to analyze the metabolites in leaves of uninfected and HLB-infected Newhall navel orange.

Smart tree crop sprayer utilizing sensor fusion and artificial intelligence →

Delivering the appropriate amount of chemicals for each plant based on their needs is essential for reducing costs, waste, and labor. Smart agricultural machinery can provide variable-rate technologies using artificial intelligence (AI) and machine vision to optimize spraying applications. In this research, a low-cost smart sensing system for controlling airblast tree crop sprayers was designed and evaluated using citrus as a case study. The prototype comprised a LiDAR, machine vision, GPS, flow meters, sensor fusion, and AI to scan trees for tree height, tree classification, and fruit counting. Specifically, this smart sensing system can detect and classify objects to tree or non-tree (e.g., human, field constructions), measure tree height and canopy density, and detect and count fruit. Based on this information, it controls spraying nozzles to optimize spraying applications. A novel software was written in C++ and ran on an Nvidia Jetson Xavier NX embedded computer to process and control the data utilizing data fusion and AI techniques. The smart sensing system's results for tree height estimate indicated a relatively low average error of 6%. A convolutional neural network (CNN) was used to perform tree classification with an average accuracy of 84% in classifying the collected imagery into mature, young, dead, and non-tree objects. The fruit count module (also a CNN) had an F1 score of 89%, compared to ground-truth labeled images of mature and immature citrus fruits. Finally, the adoption of this new sensing method reduced spraying volume by 28%, compared to traditional spraying applications.

Why it matches plant phenotyping methods樹高、樹冠密度、果実数などの植物形質をLiDAR・画像・センサ融合・AIで取得するシステムを開発し、誤差やF1スコアで技術検証しているため、散布制御への応用を超えてフェノタイピング手法が中心である。

abstracta low-cost smart sensing system for controlling airblast tree crop sprayers was designed and evaluated using citrus as a case study

Tree Crowns Segmentation and Classification in Overlapping Orchards Based on Satellite Images and Unsupervised Learning Algorithms. →

Smart agriculture is a new concept that combines agriculture and new technologies to improve the yield's quality and quantity as well as facilitate many tasks for farmers in managing orchards. An essential factor in smart agriculture is tree crown segmentation, which helps farmers automatically monitor their orchards and get information about each tree. However, one of the main problems, in this case, is when the trees are close to each other, which means that it would be difficult for the algorithm to delineate the crowns correctly. This paper used satellite images and machine learning algorithms to segment and classify trees in overlapping orchards. The data used are images from the Moroccan Mohammed VI satellite, and the study region is the OUARGHA citrus orchard located in Morocco. Our approach starts by segmenting the rows inside the parcel and finding all the trees there, getting their canopies, and classifying them by size. In general, the model inputs the parcel's image and other field measurements to classify the trees into three classes: missing/weak, normal, or big. Finally, the results are visualized in a map containing all the trees with their classes. For the results, we obtained a score of 0.93 F-measure in rows segmentation. Additionally, several field comparisons were performed to validate the classification, dozens of trees were compared and the results were very good. This paper aims to help farmers to quickly and automatically classify trees by crown size, even if there are overlapping orchards, in order to easily monitor each tree's health and understand the tree's distribution in the field.

Why it matches plant phenotyping methods衛星画像と機械学習により樹冠を分割し、個々の樹木を樹冠サイズ・生育状態で分類する手法が研究の中心で、現地比較による検証も行っているため。

abstractThis paper used satellite images and machine learning algorithms to segment and classify trees in overlapping orchards.

Evaluation of ultrasonic parameters as a non-invasive, rapid and in-field indicator of water stress in Citrus plants →

Non-Contact Resonant Ultrasound Spectroscopy (NC-RUS) has emerged as a powerful tool to determine plant water status in a non-destructive, non-invasive and rapid way. In this study, ultrasonic parameters directly obtained from experimental measurements in the field using NC-RUS - such as resonant frequency (fᵣₑₛ), velocity (v) and Q-factor - were evaluated as potential water stress indicators in Citrus plants. The experiments were carried out in two-year-old mandarin trees (Citrus clementina Hort. ex Tan. ‘Clemenules’) grown in pots in an open greenhouse where two different groups of plants were tested: a Control group (full irrigation) and a Drought Stress group (DS) whose irrigation was withdrawn during 7 days, followed by a 16 days recovery period. Soil water content, leaf water potential (Ψₗₑₐf) and the considered ultrasonic parameters were measured in the same leaves. fᵣₑₛ detected changes between control and DS at day 7 without irrigation. Conversely, v showed differences after day 3, which were statistically significant at day 7, enabling discrimination between C and DS groups. Hence, Q-factor was the ultrasonic parameter that showed statistically significant differences between C and DS groups at days 3 and 7. Consequently, Signal Intensity in Q during the drought treatment showed a similar evolution to Ψₗₑₐf, although with slightly lower performance. However, Q-factor sensitivity excels Ψₗₑₐf at each day studied. Finally, a linear correlation (R²=0.57) between Ψₗₑₐf and Q-factor of all experimental data measured in DS group plants along the drought treatment was found. In conclusion, the ultrasonic parameters obtained using NC-RUS and in particular the Q-factor, demonstrated to be potential new water stress indicators in Citrus trees, with the novelty of being non-destructive, non-invasive and rapid. Future work should explore its suitability for its use in irrigation scheduling for Citrus trees.

Why it matches plant phenotyping methodsNC-RUSによる超音波パラメータを植物の水ストレス状態の非破壊指標として評価・検証しており、表現型取得法が研究の中心である。

abstractNon-Contact Resonant Ultrasound Spectroscopy (NC-RUS) has emerged as a powerful tool to determine plant water status in a non-destructive, non-invasive and rapid way.

Canopy health, but not Candidatus Liberibacter asiaticus Ct values, are correlated with fruit yield in Huanglongbing affected sweet orange trees. →

In Florida, almost all citrus trees are infected with Huanglongbing (HLB), caused by the gram-negative, intracellular phloem limited bacteria Candidatus liberibacter asiaticus (CLas). Distinguishing between the severely and mildly sick trees is important for managing the groves and testing new HLB therapies. A mildly sick tree is one that produces higher fruit yield, compared to a severely sick tree, but measuring yields is laborious and time consuming. Here we characterized HLB affected sweet orange trees in the field in order to identify the specific traits that are correlated with the yields. We found that canopy volume, fruit detachment force (FDF) and the percentage of photosynthetically active radiation interception in the canopy (%INT) were positively correlated with fruit yields. Specifically, %INT measurements accurately distinguished between mild and severe trees in independent field trials. We could not find a difference in the Ct value between high and low producing HLB trees. Moreover, Ct values did not always agree with the number of CLas in the phloem that were visualized by transmission electron microscopy. Overall, our work identified an efficient way to distinguish between severe and mild HLB trees in Florida by measuring %INT and suggests that health of the canopy is more important for yields than the Ct value.

Why it matches plant phenotyping methodsHLB罹病樹の健康状態・重症度を、樹冠の光合成有効放射の遮断率（%INT）で判別する方法を提示し、独立圃場試験で精度を確認しているため、単なるルーチン測定を超えた表現型測定の検証研究である。

abstractSpecifically, %INT measurements accurately distinguished between mild and severe trees in independent field trials.

Detection and mapping of trees infected with citrus gummosis using UAV hyperspectral data →

Monitoring citrus diseases and pests in early stages is fundamental to ensure the efficiency of phytosanitary control and plant health. The various diseases caused by fungi, bacteria, viruses, and pests limit citrus production. Citrus gummosis disease, caused by the fungus Phytophthora spp., is the main fungal disease of citrus in Brazil. The lesions caused to the trunk and roots by Phytophthora spp. lead losses in production, foot and root rot, brown fruit rot, canopy discoloration and leaf yellowing. Remote sensing is a nondestructive detection technology, that has been used to detect phytosanitary problems in agricultural crops. Multi and hyperspectral sensors on board unmanned aerial vehicles (UAVs) have been extensively applied in agriculture. In this study, the capability for the detection of citrus gummosis was evaluated in two data sets. The first one considered hyperspectral images acquired with a 25 band sensor covering a spectral range from 500 nm to 840 nm, and the second data set was a simulated 3 band of multispectral sensor. The results indicated a better performance for the detection of citrus gummosis with the hyperspectral images than with three bands multispectral images. The high dimensionality of the hyperspectral data and the detailed spectral information allowed a more accurate classification of citrus gummosis infected plants. The classification maps were validated with field data and achieved an accuracy of 0.79 (F-score = 0.55) for the health map produced with multispectral data and an accuracy of 0.94 (F-score = 0.85) for the health map produced by the hyperspectral data.

Why it matches plant phenotyping methodsUAVハイパースペクトル画像により、柑橘樹の病害状態を検出・分類し、マルチスペクトルとの性能比較と圃場データによる検証を行っており、植物表現型（病害状態）の取得手法が中心です。

abstractIn this study, the capability for the detection of citrus gummosis was evaluated in two data sets.

Canopy Volume Extraction of Citrus reticulate Blanco cv. Shatangju Trees Using UAV Image-Based Point Cloud Deep Learning →

Automatic acquisition of the canopy volume parameters of the Citrus reticulate Blanco cv. Shatangju tree is of great significance to precision management of the orchard. This research combined the point cloud deep learning algorithm with the volume calculation algorithm to segment the canopy of the Citrus reticulate Blanco cv. Shatangju trees. The 3D (Three-Dimensional) point cloud model of a Citrus reticulate Blanco cv. Shatangju orchard was generated using UAV tilt photogrammetry images. The segmentation effects of three deep learning models, PointNet++, MinkowskiNet and FPConv, on Shatangju trees and the ground were compared. The following three volume algorithms: convex hull by slices, voxel-based method and 3D convex hull were applied to calculate the volume of Shatangju trees. Model accuracy was evaluated using the coefficient of determination (R2) and Root Mean Square Error (RMSE). The results show that the overall accuracy of the MinkowskiNet model (94.57%) is higher than the other two models, which indicates the best segmentation effect. The 3D convex hull algorithm received the highest R2 (0.8215) and the lowest RMSE (0.3186 m3) for the canopy volume calculation, which best reflects the real volume of Citrus reticulate Blanco cv. Shatangju trees. The proposed method is capable of rapid and automatic acquisition for the canopy volume of Citrus reticulate Blanco cv. Shatangju trees.

Why it matches plant phenotyping methodsUAV点群画像と深層学習・体積計算法により、樹冠体積という植物形態形質を自動抽出し、複数手法を比較検証しているため、方法が研究の中心である。

abstractAutomatic acquisition of the canopy volume parameters of the Citrus reticulate Blanco cv. Shatangju tree

Biochemical Origin of Raman-Based Diagnostics of Huanglongbing in Grapefruit Trees. →

Biotic and abiotic stresses cause substantial changes in plant biochemistry. These changes are typically revealed by high-performance liquid chromatography (HPLC) and mass spectroscopy-coupled HPLC (HPLC-MS). This information can be used to determine underlying molecular mechanisms of biotic and abiotic stresses in plants. A growing body of evidence suggests that changes in plant biochemistry can be probed by Raman spectroscopy, an emerging analytical technique that is based on inelastic light scattering. Non-invasive and non-destructive detection and identification of these changes allow for the use of Raman spectroscopy for confirmatory diagnostics of plant biotic and abiotic stresses. In this study, we couple HPLC and HPLC-MS findings on biochemical changes caused by Candidatus Liberibacter spp. ( Ca. L. asiaticus ) in citrus trees to the spectroscopic signatures of plant leaves derived by Raman spectroscopy. Our results show that Ca. L. asiaticus cause an increase in hydroxycinnamates, the precursors of lignins, and flavones, as well as a decrease in the concentration of lutein that are detected by Raman spectroscopy. These findings suggest that Ca. L. asiaticus induce a strong plant defense response that aims to exterminate bacteria present in the plant phloem. This work also suggests that Raman spectroscopy can be used to resolve stress-induced changes in plant biochemistry on the molecular level.

Why it matches plant phenotyping methodsラマン分光法を用いて植物葉の病原体誘導性の生化学変化を検出・診断し、HPLC/HPLC-MS所見と対応づけて検証しているため、植物病害状態の計測手法が中心である。

titleBiochemical Origin of Raman-Based Diagnostics of Huanglongbing in Grapefruit Trees.

Determining leaf stomatal properties in citrus trees utilizing machine vision and artificial intelligence →

Identifying and quantifying the number and size of stomata on leaf surfaces is useful for a wide range of plant ecophysiological studies, specifically those related to water-use efficiency of different plant species or agricultural crops. The time-consuming nature of manually counting and measuring stomata have limited the utility of manual methods for large-scale precision agriculture applications. A deep learning segmentation network was developed to automate the analysis of stomatal density and size and to distinguish between open and closed stomata using citrus trees grafted on different rootstocks as a model system. A novel method was developed utilizing the Mask-RCNN algorithm, which allows identification, quantification, and characterization of stomata from leaf epidermal peel microscopic images with an accuracy of up to 99%. Moreover, this method permits the differentiation of open and closed stomata with 98% precision and measurement of individual stomata size. In the citrus model system, significant differences in the size and density of stomata and diurnal regulation patterns were detected that were associated with the rootstock cultivar on which the trees were grafted. Nearly 9000 individual stomata were analyzed, which would have been impractical using manual methods. The novel automated method presented here is not only accurate, but also rapid and low-cost, and can be applied to a variety of crop and non-crop plant species.

Why it matches plant phenotyping methods柑橘葉の気孔密度・サイズ・開閉状態を画像から自動抽出するMask-RCNN手法を開発し、精度を検証しており、植物フェノタイピング手法が中心である。

abstractA deep learning segmentation network was developed to automate the analysis of stomatal density and size and to distinguish between open and closed stomata using citrus trees grafted on different rootstocks as a model system.

In situ diagnosis of mature HLB-asymptomatic citrus fruits by laser-induced breakdown spectroscopy. →

Laser-induced breakdown spectroscopy (LIBS) is a promising alternative to conventional methods in classifying citrus huanglongbing (HLB). Mature citrus fruits with similar features were picked and divided into healthy and HLB-asymptomatic groups. LIBS spectra and images were collected by focusing a laser on fresh fruit surfaces without sample preparation. The pH value and soluble solids content of juice as the indicators of acidity and sugar were detected, and the content of Ca, Zn, and K in peel and pulp was analyzed. The characteristic lines from LIBS spectra were extracted by continuous wavelet transform and principal component analysis (PCA). The t -test of these indicators displayed significant difference between the two groups. Fisher discriminant analysis and multilayer perception neural network (MLP) were applied to identify the disease. The classification accuracy reached 100% by PCA-MLP. The results show that LIBS can realize in situ detection of citrus HLB fruits.

Why it matches plant phenotyping methodsLIBSによる非破壊スペクトル・画像取得と解析を中核として、柑橘果実のHLB感染状態を識別しているため、植物病害状態のフェノタイピング手法に該当する。

abstractLIBS spectra and images were collected by focusing a laser on fresh fruit surfaces without sample preparation.

Comparison of Four Systems to Test the Tolerance of 'Fortune' Mandarin Tissue Cultured Plants to Alternaria alternata . →

Alternaria brown spot is a severe disease that affects leaves and fruits on susceptible mandarin and mandarin-like cultivars, and is produced by Alternaria alternata . Consequently, there is an urge to obtain new cultivars resistant to A. alternata , and mutation breeding together with tissue culture can help shorten the process. However, a protocol for the in vitro selection of resistant citrus genotypes is lacking. In this study, four methods to evaluate the sensitivity to Alternaria of mandarin 'Fortune' explants in in vitro culture were tested. The four tested systems consisted of: (1) the addition of the mycotoxin, produced by A. alternata in 'Fortune', to the propagation culture media, (2) the addition of the A. alternata culture filtrate to the propagation culture media, (3) the application of the mycotoxin to the intact shoot leaves, and (4) the application of the mycotoxin to the previously excised and wounded leaves. After analyzing the results, only the addition of the A. alternata culture filtrate to the culture media and the application of the mycotoxin to the wounded leaves produced symptoms of infection. However, the addition of the fungus culture filtrate to the culture media produced results, which might indicate that, in addition to the mycotoxin, many other unknown elements that can affect the plant growth and behavior could be found in the fungus culture filtrate. Therefore, the application of the toxin to the excised and wounded leaves seems to be the most reliable method to analyze sensitivity to Alternaria of 'Fortune' explants cultured in vitro.

Why it matches plant phenotyping methodsAlternaria感受性（感染症状）を評価する4つの植物表現型取得法を比較し、信頼性の高い評価プロトコルを選定しており、病害表現型の測定法が研究の中心である。

abstractIn this study, four methods to evaluate the sensitivity to Alternaria of mandarin 'Fortune' explants in in vitro culture were tested.

Evaluating the Performance of Hyperspectral Leaf Reflectance to Detect Water Stress and Estimation of Photosynthetic Capacities →

Advanced techniques capable of early, rapid, and nondestructive detection of the impacts of drought on fruit tree and the measurement of the underlying photosynthetic traits on a large scale are necessary to meet the challenges of precision farming and full prediction of yield increases. We tested the application of hyperspectral reflectance as a high-throughput phenotyping approach for early identification of water stress and rapid assessment of leaf photosynthetic traits in citrus trees by conducting a greenhouse experiment. To this end, photosynthetic CO2 assimilation rate (Pn), stomatal conductance (Cond) and transpiration rate (Trmmol) were measured with gas-exchange approaches alongside measurements of leaf hyperspectral reflectance from citrus grown across a gradient of soil drought levels six times, during 20 days of stress induction and 13 days of rewatering. Water stress caused Pn, Cond, and Trmmol rapid and continuous decline throughout the entire drought period. The upper layer was more sensitive to drought than middle and lower layers. Water stress could also bring continuous and dynamic changes of the mean spectral reflectance and absorptance over time. After trees were rewatered, these differences were not obvious. The original reflectance spectra of the four water stresses were surprisingly of low diversity and could not track drought responses, whereas specific hyperspectral spectral vegetation indices (SVIs) and absorption features or wavelength position variables presented great potential. The following machine-learning algorithms: random forest (RF), support vector machine (SVM), gradient boost (GDboost), and adaptive boosting (Adaboost) were used to develop a measure of photosynthesis from leaf reflectance spectra. The performance of four machine-learning algorithms were assessed, and RF algorithm yielded the highest predictive power for predicting photosynthetic parameters (R2 was 0.92, 0.89, and 0.88 for Pn, Cond, and Trmmol, respectively). Our results indicated that leaf hyperspectral reflectance is a reliable and stable method for monitoring water stress and yield increase, with great potential to be applied in large-scale orchards.

Why it matches plant phenotyping methods柑橘葉のハイパースペクトル反射を用いて水ストレスと光合成形質を推定する高スループット表現型計測法を適用・評価し、機械学習モデルの性能比較も行っているため、方法が中心的である。

abstractWe tested the application of hyperspectral reflectance as a high-throughput phenotyping approach for early identification of water stress and rapid assessment of leaf photosynthetic traits in citrus trees

Discrimination of Genetically Very Close Accessions of Sweet Orange ( Citrus sinensis L. Osbeck) by Laser-Induced Breakdown Spectroscopy (LIBS). →

The correct recognition of sweet orange ( Citrus sinensis L. Osbeck) variety accessions at the nursery stage of growth is a challenge for the productive sector as they do not show any difference in phenotype traits. Furthermore, there is no DNA marker able to distinguish orange accessions within a variety due to their narrow genetic trace. As different combinations of canopy and rootstock affect the uptake of elements from soil, each accession features a typical elemental concentration in the leaves. Thus, the main aim of this work was to analyze two sets of ten different accessions of very close genetic characters of three varieties of fresh citrus leaves at the nursery stage of growth by measuring the differences in elemental concentration by laser-induced breakdown spectroscopy (LIBS). The accessions were discriminated by both principal component analysis (PCA) and a classifier based on the combination of classification via regression (CVR) and partial least square regression (PLSR) models, which used the elemental concentrations measured by LIBS as input data. A correct classification of 95.1% and 80.96% was achieved, respectively, for set 1 and set 2. These results showed that LIBS is a valuable technique to discriminate among citrus accessions, which can be applied in the productive sector as an excellent cost-benefit tool in citrus breeding programs.

Why it matches plant phenotyping methodsLIBSによる葉の元素濃度測定と分類モデルを用い、近縁カンキツ系統を識別する手法を検証・適用しており、元素濃度という植物状態の取得・解析が研究の中心である。

abstractThe accessions were discriminated by both principal component analysis (PCA) and a classifier based on the combination of classification via regression (CVR) and partial least square regression (PLSR) models, which used the elemental concentrations measured by LIBS as input data.

Detection of early decay on citrus using LW-NIR hyperspectral reflectance imaging coupled with two-band ratio and improved watershed segmentation algorithm. →

Decay is a serious problem in citrus storage and transportation. However, the automatic detection of decayed citrus remains a problem. In this study, the long wavelength near-infrared (LW-NIR) hyperspectra reflectance images (1000-1850 nm) of oranges were obtained, and an effective method to detect decayed citrus was proposed. Three effective wavelength selection algorithms and two classification algorithms were used to build decay detection models in pixel-level, as well as the two-band ratio images, pseudo-color image enhancement and improved watershed segmentation were used to build decay detection models in image-level. The image-level detection method proposed in this study obtained a total success rate of 92% for all fruit, indicating its potential to detect decayed oranges online. Moreover, the LW-NIR hyperspectral reflectance imaging is verified as a useful method to detect surface defects of fruits.

Why it matches plant phenotyping methodsLW-NIRハイパースペクトル画像と画像処理による柑橘果実の腐敗・表面欠陥検出法を開発し、検出性能も評価しており、植物器官の状態を取得する方法が研究の中心である。

abstractan effective method to detect decayed citrus was proposed

Canopy Parameter Estimation of Citrus grandis var. Longanyou Based on LiDAR 3D Point Clouds →

The characteristic parameters of Citrus grandis var. Longanyou canopies are important when measuring yield and spraying pesticides. However, the feasibility of the canopy reconstruction method based on point clouds has not been confirmed with these canopies. Therefore, LiDAR point cloud data for C. grandis var. Longanyou were obtained to facilitate the management of groves of this species. Then, a cloth simulation filter and European clustering algorithm were used to realize individual canopy extraction. After calculating canopy height and width, canopy reconstruction and volume calculation were realized using six approaches: by a manual method and using five algorithms based on point clouds (convex hull, CH; convex hull by slices; voxel-based, VB; alpha-shape, AS; alpha-shape by slices, ASBS). ASBS is an innovative algorithm that combines AS with slices optimization, and can best approximate the actual canopy shape. Moreover, the CH algorithm had the shortest run time, and the R2 values of VCH, VVB, VAS, and VASBS algorithms were above 0.87. The volume with the highest accuracy was obtained from the ASBS algorithm, and the CH algorithm had the shortest computation time. In addition, a theoretical but preliminarily system suitable for the calculation of the canopy volume of C. grandis var. Longanyou was developed, which provides a theoretical reference for the efficient and accurate realization of future functional modules such as accurate plant protection, orchard obstacle avoidance, and biomass estimation.

Why it matches plant phenotyping methodsLiDAR点群から柑橘樹冠を抽出・再構成し、高さ・幅・体積を推定する手法を開発・比較検証しており、植物形質取得が研究の中心である。

abstractThen, a cloth simulation filter and European clustering algorithm were used to realize individual canopy extraction.

Recognition of Bloom/Yield in Crop Images Using Deep Learning Models for Smart Agriculture: A Review →

Precision agriculture is a crucial way to achieve greater yields by utilizing the natural deposits in a diverse environment. The yield of a crop may vary from year to year depending on the variations in climate, soil parameters and fertilizers used. Automation in the agricultural industry moderates the usage of resources and can increase the quality of food in the post-pandemic world. Agricultural robots have been developed for crop seeding, monitoring, weed control, pest management and harvesting. Physical counting of fruitlets, flowers or fruits at various phases of growth is labour intensive as well as an expensive procedure for crop yield estimation. Remote sensing technologies offer accuracy and reliability in crop yield prediction and estimation. The automation in image analysis with computer vision and deep learning models provides precise field and yield maps. In this review, it has been observed that the application of deep learning techniques has provided a better accuracy for smart farming. The crops taken for the study are fruits such as grapes, apples, citrus, tomatoes and vegetables such as sugarcane, corn, soybean, cucumber, maize, wheat. The research works which are carried out in this research paper are available as products for applications such as robot harvesting, weed detection and pest infestation. The methods which made use of conventional deep learning techniques have provided an average accuracy of 92.51%. This paper elucidates the diverse automation approaches for crop yield detection techniques with virtual analysis and classifier approaches. Technical hitches in the deep learning techniques have progressed with limitations and future investigations are also surveyed. This work highlights the machine vision and deep learning models which need to be explored for improving automated precision farming expressly during this pandemic.

Why it matches plant phenotyping methods作物画像から開花・収量を推定するコンピュータビジョン／深層学習手法を中心に扱うレビューであり、植物表現型取得・推定手法のレビューとして収録対象。

titleRecognition of Bloom/Yield in Crop Images Using Deep Learning Models for Smart Agriculture: A Review

Non-Destructive Measurement of the Internal Quality of Citrus Fruits Using a Portable NIR Device. →

The citrus industry has grown exponentially as a result of increasing demand on its consumption, giving it high standing among other fruit crops. Therefore, the citrus sector seeks rapid, easy, and non-destructive approaches to evaluate in real time and in situ the external and internal changes in physical and nutritional quality at any stage of fruit development or storage. In particular, vitamin C is among the most important micronutrients for consumers, but its measurement relies on laborious analytical methodologies. In this study, a portable near infrared spectroscopy (NIRS) sensor was used in combination with chemometrics to develop robust and accurate models to study the ripeness of several citrus fruits (oranges, lemons, clementines, tangerines, and Tahiti limes) and their vitamin C content. Ascorbic acid, dehydroascorbic acid, and total vitamin C were determined by HILIC-HPLC-UV, while soluble solids and total acidity were evaluated by standard analytical procedures. Partial least squares regression (PLSR) was used to build regression models which revealed suitable performance regarding the prediction of quality and ripeness parameters in all tested fruits. Models for ascorbic acid, dehydroascorbic acid, total vitamin C, soluble solids, total acidity, and juiciness showed Rcv2 = 0.77-0.87, Rcv2 = 0.29-0.79, Rcv2 = 0.77-0.86, Rcv2 = 0.75-0.97, Rcv2 = 0.24-0.92, and Rcv2 = 0.38-0.75, respectively. Prediction models of oranges and Tahiti limes showed good to excellent performance regarding all tested conditions. The resulting models confirmed that NIRS technology is a time- and cost-effective approach for predicting citrus fruit quality, which can easily be used by the various stakeholders from the citrus industry.

Why it matches plant phenotyping methods携帯型NIRセンサーとケモメトリクスを用いて、柑橘果実の成熟度・ビタミンC・品質形質を非破壊推定するモデルを開発・評価しており、形質取得法が研究の中心である。

abstracta portable near infrared spectroscopy (NIRS) sensor was used in combination with chemometrics to develop robust and accurate models to study the ripeness of several citrus fruits

Hyperspectral sensing of photosynthesis, stomatal conductance, and transpiration for citrus tree under drought condition →

Abstract Obtaining variation in water use and photosynthetic capacity is a promising route toward yield increases, but it is still too laborious for large-scale rapid monitoring and prediction. We tested the application of hyperspectral reflectance as a high-throughput phenotyping approach for early identification of water stress and rapid assessment of leaf photosynthetic traits in citrus trees. To this end, photosynthetic CO 2 assimilation rate ( Pn ), stomatal conductance ( Cond ) and transpiration rate ( Trmmol ) were measured with gas-exchange approaches alongside measurements of leaf hyperspectral reflectance from citrus grown across a gradient of soil drought levels. Water stress caused Pn, Cond and Trmmol rapid and continuous decreases in whole drought period. Upper layer was more sensitive to drought than middle and lower layers. Original reflectance spectra of three drought treatments were surprisingly of low diversity and could not track drought responses, whereas specific hyperspectral spectral vegetation indices (SVIs) and absorption features or wavelength position variables presented great potential. Performance of four machine learning algorithms were assessed and random forest (RF) algorithm yielded the highest predictive power for predicting photosynthetic parameters. Our results indicated that leaf hyperspectral reflectance was a reliable and stable method for monitoring water stress and yield increasing in large-scale orchards. Highlight An efficient and stable methods using hyperspectral features for early and pre-visual identification of drought and machine learning techniques for predicting photosynthetic capacity.

Why it matches plant phenotyping methods柑橘の光合成・気孔コンダクタンス・蒸散をハイパースペクトル反射と機械学習で推定する手法の開発・評価が中心であり、植物生理形質のハイスループット表現型計測に該当する。

abstractWe tested the application of hyperspectral reflectance as a high-throughput phenotyping approach for early identification of water stress and rapid assessment of leaf photosynthetic traits in citrus trees.

A CNN Approach to Simultaneously Count Plants and Detect Plantation-Rows from UAV Imagery →

In this paper, we propose a novel deep learning method based on a Convolutional Neural Network (CNN) that simultaneously detects and geolocates plantation-rows while counting its plants considering highly-dense plantation configurations. The experimental setup was evaluated in a cornfield with different growth stages and in a Citrus orchard. Both datasets characterize different plant density scenarios, locations, types of crops, sensors, and dates. A two-branch architecture was implemented in our CNN method, where the information obtained within the plantation-row is updated into the plant detection branch and retro-feed to the row branch; which are then refined by a Multi-Stage Refinement method. In the corn plantation datasets (with both growth phases, young and mature), our approach returned a mean absolute error (MAE) of 6.224 plants per image patch, a mean relative error (MRE) of 0.1038, precision and recall values of 0.856, and 0.905, respectively, and an F-measure equal to 0.876. These results were superior to the results from other deep networks (HRNet, Faster R-CNN, and RetinaNet) evaluated with the same task and dataset. For the plantation-row detection, our approach returned precision, recall, and F-measure scores of 0.913, 0.941, and 0.925, respectively. To test the robustness of our model with a different type of agriculture, we performed the same task in the citrus orchard dataset. It returned an MAE equal to 1.409 citrus-trees per patch, MRE of 0.0615, precision of 0.922, recall of 0.911, and F-measure of 0.965. For citrus plantation-row detection, our approach resulted in precision, recall, and F-measure scores equal to 0.965, 0.970, and 0.964, respectively. The proposed method achieved state-of-the-art performance for counting and geolocating plants and plant-rows in UAV images from different types of crops.

Why it matches plant phenotyping methodsUAV画像から植物数を計数し、植栽列を検出・地理的位置特定するCNN手法の開発と比較評価が中心であり、植物形態・個体数の画像ベース表現型計測に該当する。

abstractwe propose a novel deep learning method based on a Convolutional Neural Network (CNN) that simultaneously detects and geolocates plantation-rows while counting its plants

Design Plant Disease Detection System Using Deep Learning Convolutional Neural Network →

Indonesia is an agrarian country whose people mostly work in agriculture by contributing to the 3rd largest GDP. But on the other hand, the main problem in agriculture is the development of pests and diseases of crops. There are cases where there are crops that are attacked by diseases with less obvious symptoms for farmers. For example, in citrus plants that are attacked by CVPD. Initially, the citrus plant does not show too early symptoms of the disease, making it difficult to distinguish from healthy plants. Based on these problems early detection and identification of plant diseases are the main factors to prevent and reduce the spread of plant diseases. The study used deep learning methods with the Convolutional Neural Network (CNN) algorithm model. The dataset used comes from PlantVillage with a total of 20,639 leaf image files that have been classified based on their respective classes. The design of the model architecture is done by designing the CNN model following the DenseNet121 architecture, by changing the parameters to improve the accuracy results. Image size is 64, train shape (20639, 64, 64, 3), epoch value 50,100, and 150. The number of input layers used is 4 layers with shapes (64, 64, 3). Densenet121 shape (1024), global average pooling2D shape (1024), batch normalization 2 (1024), dropout (1024), dense (256), batch normalization 3 (256), root (Dense) (15). This research was conducted with 3 epoch iteration tests to find the best accuracy value. The training data for epoch 50,100, and 150 produces an average model accuracy of 99.38% and the average value of the model loss is 0.019% can also be seen from the testing data results for epoch 50,100, and 150 has an average model of 95.16% and can be seen also from the average value for the loss is 0.20%. Based on the algorithm that applied the resulting training accuracy of 99.58% and the accuracy of testing 96.41% then design this application is useful to accurately detect diseases in plants by using leaf imagery of the plant.

Why it matches plant phenotyping methods植物葉画像から病害状態を推定するCNNモデルの設計・精度評価が研究の中心であり、植物病害フェノタイピング手法に該当する。

abstractThe study used deep learning methods with the Convolutional Neural Network (CNN) algorithm model.

Identification of citrus immune regulators involved in defence against Huanglongbing using a new functional screening system. →

Huanglongbing (HLB) is the most devastating citrus disease in the world. Almost all commercial citrus varieties are susceptible to the causal bacterium, Candidatus Liberibacter asiaticus (CLas), which is transmitted by the Asian citrus psyllid (ACP). Currently, there are no effective management strategies to control HLB. HLB-tolerant traits have been reported in some citrus relatives and citrus hybrids, which offer a direct pathway for discovering natural defence regulators to combat HLB. Through comparative analysis of small RNA profiles and target gene expression between an HLB-tolerant citrus hybrid (Poncirus trifoliata × Citrus reticulata) and a susceptible citrus variety, we identified a panel of candidate defence regulators for HLB-tolerance. These regulators display similar expression patterns in another HLB-tolerant citrus relative, with a distinct genetic and geographic background, the Sydney hybrid (Microcitrus virgata). Because the functional validation of candidate regulators in tree crops is always challenging, we developed a novel rapid functional screening method, using a C. Liberibacter solanacearum (CLso)/potato psyllid/Nicotiana benthamiana interaction system to mimic the natural transmission and infection circuit of the HLB complex. When combined with efficient virus-induced gene silencing in N. benthamiana, this innovative and cost-effective screening method allows for rapid identification and functional characterization of regulators involved in plant immune responses against HLB, such as the positive regulator BRCA1-Associated Protein, and the negative regulator Vascular Associated Death Protein.

Why it matches plant phenotyping methods植物免疫応答を評価する新規かつ迅速な機能スクリーニング法の開発が研究の中心であり、HLB関連の植物状態・防御応答を表現型として評価している。

abstractwe developed a novel rapid functional screening method

Canine Olfactory Detection of a Non-Systemic Phytobacterial Citrus Pathogen of International Quarantine Significance. →

For millennia humans have benefitted from application of the acute canine sense of smell to hunt, track and find targets of importance. In this report, canines were evaluated for their ability to detect the severe exotic phytobacterial arboreal pathogen Xanthomonas citri pv. citri (Xcc), which is the causal agent of Asiatic citrus canker (Acc). Since Xcc causes only local lesions, infections are non-systemic, limiting the use of serological and molecular diagnostic tools for field-level detection. This necessitates reliance on human visual surveys for Acc symptoms, which is highly inefficient at low disease incidence, and thus for early detection. In simulated orchards the overall combined performance metrics for a pair of canines were 0.9856, 0.9974, 0.9257 and 0.9970, for sensitivity, specificity, precision, and accuracy, respectively, with 1-2 s/tree detection time. Detection of trace Xcc infections on commercial packinghouse fruit resulted in 0.7313, 0.9947, 0.8750, and 0.9821 for the same performance metrics across a range of cartons with 0-10% Xcc-infected fruit despite the noisy, hot and potentially distracting environment. In orchards, the sensitivity of canines increased with lesion incidence, whereas the specificity and overall accuracy was >0.99 across all incidence levels; i.e., false positive rates were uniformly low. Canines also alerted to a range of 1-12-week-old infections with equal accuracy. When trained to either Xcc-infected trees or Xcc axenic cultures, canines inherently detected the homologous and heterologous targets, suggesting they can detect Xcc directly rather than only volatiles produced by the host following infection. Canines were able to detect the Xcc scent signature at very low concentrations (10,000× less than 1 bacterial cell per sample), which implies that the scent signature is composed of bacterial cell volatile organic compound constituents or exudates that occur at concentrations many fold that of the bacterial cells. The results imply that canines can be trained as viable early detectors of Xcc and deployed across citrus orchards, packinghouses, and nurseries.

Why it matches plant phenotyping methodsイヌによる柑橘植物の病原体感染・病徴の早期検出法を開発・評価し、感度、特異度、精度、検出時間を検証しているため、植物病害状態の表現型取得法が中心である。

abstractIn simulated orchards the overall combined performance metrics for a pair of canines were 0.9856, 0.9974, 0.9257 and 0.9970, for sensitivity, specificity, precision, and accuracy, respectively, with 1-2 s/tree detection time.

3D RECONSTRUCTION OF CITRUS TREES USING AN OMNIDIRECTIONAL OPTICAL SYSTEM →

Abstract. This paper presents a feasibility study on the use of omnidirectional systems for 3D modelling of agricultural crops, aiming a systematic monitoring. Omnidirectional systems with multiple sensors have been widely used in close-range photogrammetry (CRP), which can be a good alternative to provide data for digital agriculture management. The GoPro Fusion dual-camera is the omnidirectional system used in this work. This system is composed of two cameras with fisheye lenses that cover more than 180° each one in back-to-back position. System calibration, camera orientation and 3D reconstruction of an agricultural cultivated area were performed in Agisoft Metashape software. A 360° calibration field based on coded targets (CTs) from Agisoft Metashape software was used to calibrate the omnidirectional system. The 3D reconstruction of an orange orchard was performed using fisheye images taken with GoPro Fusion. The results show the potential of using an omnidirectional system for 3D modelling in agricultural crops, in particular citrus trees. Interior orientation parameters (IOPs) was estimated using Agisoft Metashape target/software with a precision of 9 mm. A 3D reconstruction model of the orange orchard area was obtained with an accuracy of 3.8 cm, which can be considered acceptable for agricultural purposes.

Why it matches plant phenotyping methods全方位カメラの校正と3D再構成精度を評価し、柑橘樹の3Dモデル化を行う画像ベースの植物計測手法が研究の中心である。

abstractThis paper presents a feasibility study on the use of omnidirectional systems for 3D modelling of agricultural crops, aiming a systematic monitoring.

The application of artificial neural networks in modeling and predicting the effects of melatonin on morphological responses of citrus to drought stress. →

Drought stress as one of the most devastating abiotic stresses affects agricultural and horticultural productivity in many parts of the world. The application of melatonin can be considered as a promising approach for alleviating the negative impact of drought stress. Modeling of morphological responses to drought stress can be helpful to predict the optimal condition for improving plant productivity. The objective of the current study is modeling and predicting morphological responses (leaf length, number of leaves/plants, crown diameter, plant height, and internode length) of citrus to drought stress, based on four input variables including melatonin concentrations, days after applying treatments, citrus species, and level of drought stress, using different Artificial Neural Networks (ANNs) including Generalized Regression Neural Network (GRNN), Radial basis function (RBF), and Multilayer Perceptron (MLP). The results indicated a higher accuracy of GRNN as compared to RBF and MLP. The great accordance between the experimental and predicted data of morphological responses for both training and testing processes support the excellent efficiency of developed GRNN models. Also, GRNN was connected to Non-dominated Sorting Genetic Algorithm-II (NSGA-II) to optimize input variables for obtaining the best morphological responses. Generally, the validation experiment showed that ANN-NSGA-II can be considered as a promising and reliable computational tool for studying and predicting plant morphological and physiological responses to drought stress.

Why it matches plant phenotyping methods植物の形態形質を予測するANNモデルを開発・比較し、検証実験と最適化まで行っており、形質推定の計算手法が研究の中心である。

abstractThe objective of the current study is modeling and predicting morphological responses (leaf length, number of leaves/plants, crown diameter, plant height, and internode length) of citrus to drought stress

Active thermal imaging for immature citrus fruit detection →

Yield mapping for citrus fruit is a challenging task due to factors such as varying illumination conditions, clustering, and occlusions. Mapping the yield for immature citrus fruit presents an additional challenge that is the colours of fruit and leaves are almost identical. Commonly used machine vision techniques using colour cameras become less effective for immature citrus fruit detection. This study explores a novel active thermal imaging method to tackle the problem of colour similarity between immature citrus fruit and leaves. In this study, a thermal camera was combined with a water spray system that applied water mist to citrus trees. The water mist caused temperatures of both the fruit and leaf surfaces to change but at different rates. Multiple parameters of the spray system were experimented with the goal to induce as much temperature differences as possible between fruit and leaf surfaces. The combined system was tested in a citrus grove for fruit detections. Deep learning models were built based on the active thermal imaging system and tracking and fruit counting algorithms were created to count fruit in thermal videos. A mean average precision of 87.2% was achieved by the models and an accuracy of 96% was achieved when comparing the number of fruit counted by the algorithms with the true number of fruit counted manually in the field.

Why it matches plant phenotyping methods未熟果実を対象に、散水と熱画像を組み合わせた検出・計数手法を開発し、現地で性能検証している。果実数という植物器官の形質取得が中心である。

abstractThis study explores a novel active thermal imaging method to tackle the problem of colour similarity between immature citrus fruit and leaves.

Comparison of convolution neural networks for smartphone image based real time classification of citrus leaf disease →

Experts help farmers to diagnose the citrus diseases by using agriculture laboratories or viewing the visual symptoms. These methods may not be accessible to all the farmers due to the expert’s cost and non-availability of laboratories. The proposed work presents the comparison of two different Convolutional Neural Network (CNN) architectures to classify diseases of the citrus leaf. In this paper, two types of CNN architectures, such as MobileNet and Self-Structured (SSCNN) classifiers were used to detect and classify citrus leaf diseases at the vegetative stage. The proposed work prepared a smartphone image based citrus disease dataset. Both the models were trained and tested on the same citrus dataset. The performances of the models were evaluated using the accuracy and loss of the training and validation sets, respectively. The best training accuracy of the MobileNet CNN was 98% with 92% validation accuracy at the epoch 10. But the best training accuracy of the SSCNN was 98% with 99% validation accuracy at the epoch 12. The proposed system indicates that the SSCNN is more helpful and accurate for smartphone image based citrus leaf disease classification. In addition, the SSCNN algorithm takes less computation time as compared to MobileNet and it can be considered a cost-effective method for citrus disease detection.

Why it matches plant phenotyping methodsスマートフォン画像から柑橘葉の病害状態を分類するCNN手法を比較・評価し、データセットも構築しているため、植物病害表現型の取得・推定方法が中心である。

abstractThe proposed work presents the comparison of two different Convolutional Neural Network (CNN) architectures to classify diseases of the citrus leaf.

Characterization of Volatile Organic Compounds of Healthy and Huanglongbing-Infected Navel Orange and Pomelo Leaves by HS-GC-IMS. →

The Asian citrus psyllid (ACP), Diaphorina citri Kuwayama, is the only natural vector of bacteria responsible for Huanglongbing (HLB), a worldwide destructive disease of citrus. ACP reproduces and develops only on the young leaves of its rutaceous host plants. Olfactory stimuli emitted by young leaves may play an important role in ACP control and HLB detection. In this study, volatile organic compounds (VOCs) from healthy and HLB-infected young leaves of navel orange and pomelo were analyzed by headspace-gas chromatography-ion mobility spectrometry (HS-GC-IMS). A total of 36 compounds (including dimers or polymers) were identified and quantified from orange and 10 from pomelo leaves. Some compounds showed significant differences in signal intensity between healthy and HLB-infected leaves and may constitute possible indicators for HLB infection. Principal component analysis (PCA) clearly discriminated healthy and HLB-infected leaves in both orange and pomelo. HS-GC-IMS was an effective method to identify VOCs from leaves. This study may help develop new methods for detection of HLB or find new attractants or repellents of ACP for prevention of HLB.

Why it matches plant phenotyping methodsHS-GC-IMSによる葉の揮発性成分測定とPCA分類を用いて、HLB感染状態の検出可能性を評価しており、植物の病害状態を抽出する測定手法が中心です。

abstractSome compounds showed significant differences in signal intensity between healthy and HLB-infected leaves and may constitute possible indicators for HLB infection.

Racing against stomatal attenuation: rapid CO2 response curves more reliably estimate photosynthetic capacity than steady state curves in a low conductance species →

A/Ci curves are an important gas-exchange-based approach to understanding the regulation of photosynthesis, describing the response of net CO2 assimilation (A) to leaf internal concentration of CO2 (Ci). Low stomatal conductance species pose a challenge to the measurement of A/Ci curves by reducing the signal-to-noise ratio of gas exchange measures. Additionally, the stomatal attenuation effect of elevated ambient CO2 leads to further reduction of conductance and may lead to erroneous interpretation of high Ci responses of A. Rapid A/Ci response (RACiR) curves offer a potential practice to develop A/Ci curves faster than the stomatal closure response to elevated CO2. We used the moderately low conductance Citrus to compare traditional steady state (SS) A/Ci curves with RACiR curves. SS curves failed more often than RACiR curves. Overall parameter estimates were the same between SS and RACiR curves. When low stomatal conductance values were removed, triose-phosphate utilization (TPU) limitation estimates increased. Overall RACiR stomatal conductance values began and remained higher than SS values. Based on the comparable resulting parameter estimates, higher likelihood of success and reduced measurement time, we propose RACiR as a valuable tool to measure A/Ci responses in low conductance species.

Why it matches plant phenotyping methods低コンダクタンス植物の光合成能力を測定するA/Ciガス交換法について、従来法との比較検証と測定手順の改善を中心に扱っているため、植物フェノタイピング手法として含める。

abstractRapid A/Ci response (RACiR) curves offer a potential practice to develop A/Ci curves faster than the stomatal closure response to elevated CO2.

Detection of Citrus Huanglongbing Based on Multi-Input Neural Network Model of UAV Hyperspectral Remote Sensing →

Citrus is an important cash crop in the world, and huanglongbing (HLB) is a destructive disease in the citrus industry. To efficiently detect the degree of HLB stress on large-scale orchard citrus trees, an UAV (Uncrewed Aerial Vehicle) hyperspectral remote sensing tool is used for HLB rapid detection. A Cubert S185 (Airborne Hyperspectral camera) was mounted on the UAV of DJI Matrice 600 Pro to capture the hyperspectral remote sensing images; and a ASD Handheld2 (spectrometer) was used to verify the effectiveness of the remote sensing data. Correlation-proven UAV hyperspectral remote sensing data were used, and canopy spectral samples based on single pixels were extracted for processing and analysis. The feature bands extracted by the genetic algorithm (GA) of the improved selection operator were 468 nm, 504 nm, 512 nm, 516 nm, 528 nm, 536 nm, 632 nm, 680 nm, 688 nm, and 852 nm for the HLB detection. The proposed HLB detection methods (based on the multi-feature fusion of vegetation index) and canopy spectral feature parameters constructed (based on the feature band in stacked autoencoder (SAE) neural network) have a classification accuracy of 99.33% and a loss of 0.0783 for the training set, and a classification accuracy of 99.72% and a loss of 0.0585 for the validation set. This performance is higher than that based on the full-band AutoEncoder neural network. The field-testing results show that the model could effectively detect the HLB plants and output the distribution of the disease in the canopy, thus judging the plant disease level in a large area efficiently.

Why it matches plant phenotyping methodsUAVハイパースペクトル画像から柑橘樹のHLB症状・病害レベルを推定するセンシングおよびニューラルネットワーク手法が中心であり、技術検証も実施している。

abstractan UAV (Uncrewed Aerial Vehicle) hyperspectral remote sensing tool is used for HLB rapid detection

Mass spectrometry imaging as a potential technique for diagnostic of Huanglongbing disease using fast and simple sample preparation. →

Huanglongbing (HLB) is a disease of worldwide incidence that affects orange trees, among other commercial varieties, implicating in great losses to the citrus industry. The disease is transmitted through Diaphorina citri vector, which inoculates Candidatus Liberibacter spp. in the plant sap. HLB disease lead to blotchy mottle and fruit deformation, among other characteristic symptoms, which induce fruit drop and affect negatively the juice quality. Nowadays, the disease is controlled by eradication of sick, symptomatic plants, coupled with psyllid control. Polymerase chain reaction (PCR) is the technique most used to diagnose the disease; however, this methodology involves high cost and extensive sample preparation. Mass spectrometry imaging (MSI) technique is a fast and easily handled sample analysis that, in the case of Huanglongbing allows the detection of increased concentration of metabolites associated to the disease, including quinic acid, phenylalanine, nobiletin and sucrose. The metabolites abieta-8,11,13-trien-18-oic acid, suggested by global natural product social molecular networking (GNPS) analysis, and 4-acetyl-1-methylcyclohexene showed a higher distribution in symptomatic leaves and have been directly associated to HLB disease. Desorption electrospray ionization coupled to mass spectrometry imaging (DESI-MSI) allows the rapid and efficient detection of biomarkers in sweet oranges infected with Candidatus Liberibacter asiaticus and can be developed into a real-time, fast-diagnostic technique.

Why it matches plant phenotyping methodsDESI-MSIを用いて感染植物の症状関連代謝物を空間的に検出し、HLB病の迅速診断法として評価・提案しており、植物の病害状態の取得が中心的な方法論的貢献である。

abstractMass spectrometry imaging (MSI) technique is a fast and easily handled sample analysis that, in the case of Huanglongbing allows the detection of increased concentration of metabolites associated to the disease

Monitoring the vegetation vigor in heterogeneous citrus and olive orchards. A multiscale object-based approach to extract trees’ crowns from UAV multispectral imagery →

Precision agriculture (PA) constitutes one of the most critical sectors of remote sensing applications that allow obtaining spatial segmentation and within-field variability information from field crops. In the last decade, an increasing source of information is provided by unmanned aerial vehicle (UAVs) platforms, mainly equipped with optical multispectral cameras, to map, monitor, and analyze, temporal and spatial variations of vegetation using ad hoc spectral vegetation indices (VIs). Considering the centimeter or sub-centimeter spatial resolution of UAV imagery, the geographic object-based image analysis (GEOBIA) approach, is becoming prevalent in UAV remote sensing applications. In the present paper, we propose a quick and reliable semi-automatic workflow implemented to process multispectral UAV imagery and aimed at the detection and extraction of olive and citrus trees’ crowns to obtain vigor maps in the framework of PA. We focused our attention on the choice of GEOBIA data input and parameters, taking into consideration its replicability and reliability in the case of heterogeneous tree orchards. The heterogeneity concerns the different tree plantation distances and composition, different crop management (irrigation, pruning, weeding), and different tree age, height, and crown diameters. The proposed GEOBIA workflow was implemented in the eCognition Developer 9.5, coupling the use of multispectral and topographic information surveyed using the Tetracam µ-MCA06 snap multispectral camera at 4 cm of ground sample distance (GSD). Three different study sites in heterogeneous citrus (Bergamot and Clementine) and olive orchards located in the Calabria region (Italy) were provided. Multiresolution segmentation was implemented using spectral and topographic band layers and optimized by applying a trial-and-error approach. The classification step was implemented as process-tree and based on a rule set algorithm, therefore easily adaptable and replicable to other datasets. Decision variables for image classification were spectral vegetation indices (NDVI, SAVI, CVI) and topographic layers (DSM and CHM). Vigor maps were based on NDVI and NDRE and allowed to highlight those areas with low vegetative vigor. The accuracy assessment was based on a per-pixel approach and computed through the F-score (F). The obtained results are promising, considering that the resulting accuracy was high, with F-score ranging from 0.85 to 0.91 for olive and bergamot, respectively. Our proposed workflow, which has proved effective in datasets of different complexity, finds its strong point is the speed of execution and on its repeatability to other different crops with few adjustments. It appears worth of interest to highlights that it requests a working day of two good skilled operators in geomatics and computer image processing, from the on-field data collection to the obtaining of vigor maps.

Why it matches plant phenotyping methodsUAVマルチスペクトル画像から樹冠を抽出し、植生 vigor を推定する再現可能な画像解析ワークフローを開発・精度評価しており、植物表現型取得が研究の中心である。

abstractwe propose a quick and reliable semi-automatic workflow implemented to process multispectral UAV imagery and aimed at the detection and extraction of olive and citrus trees’ crowns to obtain vigor maps

Integrated detection of citrus fruits and branches using a convolutional neural network →

The key technology for a fruit picking robot is to identify fruits in different occlusion states. Based on the mask regional convolutional neural network (Mask R-CNN) and a branch segment merging algorithm, an integrated system was developed to simultaneously detect and measure citrus fruits and branches. A training dataset was constructed for fruit and tree appearance, including single fruit, multiple fruits, occluded fruits, branches and trunk. A segmental labeling method for random and irregular branches is proposed to improve the precision of the Mask R-CNN. Based on the segmental mask regions identified by this model, a more precise bounding box is obtained by calculating the minimum enclosing rectangle of mask regions. Then, a branch segment merging algorithm reconstructs branches and the trunk. Diameters of fruits and branches are obtained by mapping the color image onto the depth image. The average precision of fruit and branch recognition are 88.15% and 96.27%, respectively. The average measurement error of fruits’ transverse diameters, fruits’ longitudinal diameters, and branch diameters are 2.52, 2.29, and 1.17 mm, respectively. Experiments show the detection system has good performance for all types of fruits and occlusions. This vision system can effectively help the robot to plan the appropriate picking path and avoid obstacles.

Why it matches plant phenotyping methodsMask R-CNNと深度画像を用いて果実・枝を検出し、直径を定量する画像ベースの植物形質計測システムが中心であり、単なる収穫対象の位置検出を超えている。

abstractan integrated system was developed to simultaneously detect and measure citrus fruits and branches.

Assessing the Orange Tree Crown Volumes Using Google Maps as a Low-Cost Photogrammetric Alternative →

The accurate assessment of tree crowns is important for agriculture, for example, to adjust spraying rates, to adjust irrigation rates or even to estimate biomass. Among the available methodologies, there are the traditional methods that estimate with a three-dimensional approximation figure, the HDS (High Definition Survey), or TLS (Terrestrial Laser Scanning) based on LiDAR technology, the aerial photogrammetry that has re-emerged with unmanned aerial vehicles (UAVs), as they are considered low cost. There are situations where either the cost or location does not allow for modern methods and prices such as HDS or the use of UAVs. This study proposes, as an alternative methodology, the evaluation of images extracted from Google Maps (GM) for the calculation of tree crown volume. For this purpose, measurements were taken on orange trees in the south of Spain using the four methods mentioned above to evaluate the suitability, accuracy, and limitations of GM. Using the HDS method as a reference, the photogrammetric method with UAV images has shown an average error of 10%, GM has obtained approximately 50%, while the traditional methods, in our case considering ellipsoids, have obtained 100% error. Therefore, the results with GM are encouraging and open new perspectives for the estimation of tree crown volumes at low cost compared to HDS, and without geographical flight restrictions like those of UAVs.

Why it matches plant phenotyping methodsオレンジ樹冠体積という植物形態形質を、Google Maps画像を用いて推定する低コスト測定法として提案・比較検証しており、表現型取得法が研究の中心である。

abstractThis study proposes, as an alternative methodology, the evaluation of images extracted from Google Maps (GM) for the calculation of tree crown volume.

Raman Spectroscopy vs Quantitative Polymerase Chain Reaction In Early Stage Huanglongbing Diagnostics. →

Raman spectroscopy (RS) is an emerging analytical technique that can be used to develop and deploy precision agriculture. RS allows for confirmatory diagnostic of biotic and abiotic stresses on plants. Specifically, RS can be used for Huanglongbing (HLB) diagnostics on both orange and grapefruit trees, as well as detection and identification of various fungal and viral diseases. The questions that remain to be answered is how early can RS detect and identify the disease and whether RS is more sensitive than qPCR, the "golden standard" in pathogen diagnostics? Using RS and HLB as case study, we monitored healthy (qPCR-negative) in-field grown citrus trees and compared their spectra to the spectra collected from healthy orange and grapefruit trees grown in a greenhouse with restricted insect access and confirmed as HLB free by qPCR. Our result indicated that RS was capable of early prediction of HLB and that nearly all in-field qPCR-negative plants were infected by the disease. Using advanced multivariate statistical analysis, we also showed that qPCR-negative plants exhibited HLB-specific spectral characteristics that can be distinguished from unrelated nutrition deficit characteristics. These results demonstrate that RS is capable of much more sensitive diagnostics of HLB compared to qPCR.

Why it matches plant phenotyping methodsRaman分光法を用いた植物病害（HLB）の早期診断を中心に、qPCRとの比較および多変量解析による技術評価を行っているため、植物フェノタイピング手法として収載する。

abstractUsing RS and HLB as case study, we monitored healthy (qPCR-negative) in-field grown citrus trees and compared their spectra to the spectra collected from healthy orange and grapefruit trees grown in a greenhouse with restricted insect access and confirmed as HLB free by qPCR.

Agroview: Cloud-based application to process, analyze and visualize UAV-collected data for precision agriculture applications utilizing artificial intelligence →

Traditional sensing technologies in specialty crops production, for pest and disease detection and field phenotyping, rely on manual sampling and are time consuming and labor intensive. Since availability of personnel trained for field scouting is a major problem, small Unmanned Aerial Vehicles (UAVs) equipped with various sensors can simplify the surveying procedure, decrease data collection time, and reduce cost. To accurate and rapidly process, analyze and visualize data collected from UAVs and other platforms (e.g. small airplanes, satellites, ground platforms), a cloud and artificial intelligence (AI) based application (named Agroview) was developed. This interactive and user-friendly application can: (i) detect, count and geo-locate plants and plant gaps (locations with dead or no plants); (ii) measure plant height and canopy size (plant inventory); (iii) develop plant health (or stress) maps. In this study, the use of this Agroview application to evaluate phenotypic characteristics of citrus trees (as a case study) is presented. It was found, that this emerging technology detected citrus trees with mean absolute percentage error (MAPE) of 2.3% in a commercial citrus orchard with 175,977 trees (1,871 acres; 39 normal and high-density spacing blocks). Furthermore, it accurately estimated tree height with 4.5% and 12.93% MAPE for normal and high-density spacing respectively, and canopy size with MAPE of 12.9% and 34.6% for normal and high-density spacing respectively. It provides a consistent, more direct, cost-effective and rapid method for field survey and plant phenotyping.

Why it matches plant phenotyping methodsUAV画像を処理するクラウド型プラットフォームを開発し、樹木の検出、樹高、樹冠サイズ、健康状態を定量化・検証しており、植物表現型取得が研究の中心です。

abstracta cloud and artificial intelligence (AI) based application (named Agroview) was developed.

Laser-induced fluorescence spectroscopy for early disease detection in grapefruit plants. →

Biotic and abiotic stress both cause a considerable decrease in the chlorophyll content in plant leaves, which provides a means for the early diagnosis of diseases in plants. The emergence of diseases affects the fluorescence of phenolic compounds and chlorophyll, which have emissions located at 530, 686 and 735 nm. Herein, it was found that the intensity of the emission band of phenolic compounds at 530 nm increased and that of chlorophyll at 735 nm decreased with the onset of diseases. Statistical analysis through principal component analysis (PCA) and partial least squares regression (PLSR) was performed, which differentiated between apparently healthy leaf sites and diseased leaves, providing a basis for the detection of diseases in the early stages. The PLSR model was validated through the coefficient of determination (R 2 ), standard error of prediction (SEP) and standard error of calibration (SEC) with the values of 0.99, 0.394 and 0.0.401, respectively, which authenticated the model. The prediction accuracy of the model was evaluated through root mean square error in prediction (RMSEP), with a value of 0.14, by predicting 22 unknown emission spectra of different leaf sites. Both the PCA and PLSR models produced similar results, proving that fluorescence spectroscopy is an excellent tool for early disease detection in plants.

Why it matches plant phenotyping methods植物葉の蛍光スペクトルから病害状態を検出し、PCA/PLSRモデルを検証しているため、植物フェノタイピング手法が中心です。

titleLaser-induced fluorescence spectroscopy for early disease detection in grapefruit plants.

Deep learning techniques for estimation of the yield and size of citrus fruits using a UAV →

Accurate and early estimation of citrus yields is important for both producers and agricultural cooperatives to be competitive and make informed decisions when selling their products. Yield estimation is key for predicting stock volumes, avoiding stock ruptures and planning harvesting operations. Visual yield estimations have traditionally been employed, resulting in inaccurate and misleading information. The main goal of this study was to develop an automated image processing methodology to detect, count and estimate the size of citrus fruits on individual trees using deep learning techniques. During 3 consecutive annual campaigns, a total of 20 trees from a commercial citrus grove were monitored using images captured from an unmanned aerial vehicle (UAV). These trees were harvested manually, and fruit sizes were measured. A Faster R-CNN Deep Learning model was trained using a custom dataset to detect oranges in the obtained images. An average standard error (SE) of 6.59 % was obtained between visual counting and the model’s fruit detection. Using the detected fruits, fruit size estimation was also performed. The promising results obtained indicate that this size estimation method can be employed for size discrimination prior to harvest. A model based on Long Short-term Memory (LSTM) was trained for yield estimation per tree and for a total yield estimation. The actual and estimated yields per tree were compared, resulting in an approximate error of SE = 4.53 % and a standard deviation of SD = 0.97 Kg. The actual total yield, the estimated total yield and the total yield estimated by an expert technician were compared. The error in the estimation by the technician was SE = 13.74 %, while the errors in the model were SE = 7.22 % and SD = 4083.58 Kg. These promising results demonstrate the potential of the present technique to provide yield estimates for citrus fruits or even other types of fruit.

Why it matches plant phenotyping methodsUAV画像と深層学習により、個体樹ごとの果実検出・計数・サイズおよび収量を推定する方法を開発・検証しており、植物形質取得が研究の中心である。

abstractThe main goal of this study was to develop an automated image processing methodology to detect, count and estimate the size of citrus fruits on individual trees using deep learning techniques.

Comparison of machine learning methods for citrus greening detection on UAV multispectral images →

Citrus Huanglongbing (HLB), also known as citrus greening, is the most destructive disease in the citrus industry. Detecting this disease as early as possible and eradicating the roots of HLB-infected trees can control its spread. Ground diagnosis is time-consuming and laborious. Large area monitoring method of citrus orchard with high accuracy is rare. This study evaluates the feasibility of large area detection of citrus HLB by low altitude remote sensing and commits to improve the accuracy of large-area detection. A commercial multispectral camera (ADC-lite) mounted on DJI M100 UAV(unmanned Aerial Vehicle) was used to collect green, red and near-infrared multispectral image of large area citrus orchard, a linear-stretch was performed to remove noise pixel, vegetation indices (VIs) were calculated followed by correlation analysis and feature compression using PCA (principal components analysis) and AutoEncoder to discover potential features. Several machine learning algorithms, such as support vector machine (SVM), k-nearest neighbour (kNN), logistic regression (LR), naive Bayes and ensemble learning, were compared to model the healthy and HLB-infected samples after parameter optimization. The results showed that the feature of PCA features of VIs combining with original DN (digital numbers) value generally have highest accuracy and agreement in all models, and the ensemble learning and neural network approaches had strong robustness and the best classification results (100% in AdaBoost and 97.28% in neural network) using threshold strategy.

Why it matches plant phenotyping methodsUAVマルチスペクトル画像と複数の機械学習手法を比較し、柑橘樹のHLB感染状態を大規模に推定する方法が研究の中心であるため、植物病害フェノタイピング手法として採用する。

abstractThis study evaluates the feasibility of large area detection of citrus HLB by low altitude remote sensing and commits to improve the accuracy of large-area detection.

Machine Learning-based inference system to detect the phenological stage of a citrus crop for helping deficit irrigation techniques to be automatically applied. →

This paper presents a system that helps farmers to irrigate crops, minimizing water consumption, while productivity is kept, when deficit irrigation techniques are applied, according to the phenological stage of such crop. Such stage is automatically inferred by using a Machine Learning-based technique, which uses single images, which can be acquired by simply using a low cost commercial camera (even the one embedded in a smartphone), as inputs. Specifically, this work compares several Machine Learning approaches, in particular, classical and deep neural networks trained with a dataset obtained from taking multiple real images from a citrus crop. Such images represent different growing stages of the citrus associated to different phenological stages. Since, according to the deficit irrigation approach, the amount of water that can be reduced without affecting the yield depends on the phenological stage of the crop, once such stage is inferred, a Decision Support System uses such information for automatically programming irrigation. The paper also remarks the main advantages of using a single camera as unique sensor in terms of low economic cost as opposed to other systems that uses more expensive and invasive sensors in the crop. In addition, as a smartphone camera could be used as sensor, the smartphone itself could be used as computing device to run the phenological stage detector in real time, and to interact with the Decision Support System by using Cloud and Edge computing technologies. Finally, a set of experiments show the main results obtained after testing different Machine Learning approaches. After comparing such approaches, the best choice is selected to be integrated as a part of the mentioned Decision Support System.

Why it matches plant phenotyping methods画像から柑橘の生育・フェノロジー段階を推定する機械学習手法を開発・比較評価しており、植物表現型の取得方法が中心的です。

abstractSuch stage is automatically inferred by using a Machine Learning-based technique, which uses single images

A new visible band index (vNDVI) for estimating NDVI values on RGB images utilizing genetic algorithms →

Several vegetation indices have been developed, with the normalized difference vegetation index (NDVI) been the most studied and commonly used. To generate an NDVI map, a relatively high-cost multispectral sensor is required; but currently, most UAVs are equipped with low-cost RGB cameras. For that reason, other indices that utilize RGB data have been developed to generate maps similar to NDVI and minimize the data acquisition cost, such as the triangular greenness index (TGI) and the visible atmospheric resistant index (VARI). However, several studies found that these indices cannot be recommended as reliable general-purpose crop health indicators. This study utilizes a genetic algorithm to develop a new visible index (visible NDVI; vNDVI) that estimates NDVI values of vegetation from uncalibrated RGB cameras mounted on UAVs (or other remote sensing platforms). Three experiments were conducted to create and validate the proposed index. First, the NDVI values generated from a multispectral camera were compared with the NDVI values generated by a hyperspectral camera. In the second experiment, the vNDVI formula was created using a genetic algorithm. The third experiment validates the proposed vNDVI, generated from two uncalibrated RGB cameras, in three different crops (citrus, grapes, and sugarcane). The proposed vNDVI proved to be highly accurate on estimating NDVI values by just using RGB cameras, with an overall mean percentage error of 6.89% and a mean average error of 0.052 in all three crops, providing a low-cost alternative for remote sensing and plant phenotyping.

Why it matches plant phenotyping methodsRGB画像から植物のNDVIを推定する新規可視指数を遺伝的アルゴリズムで開発し、複数作物・カメラで検証しており、植物表現型取得手法が中心である。

abstractThis study utilizes a genetic algorithm to develop a new visible index (visible NDVI; vNDVI) that estimates NDVI values of vegetation from uncalibrated RGB cameras mounted on UAVs (or other remote sensing platforms).

Swarm Intelligence Based Detection of Citrus Plant Diseases and Their Severity Level →

The quality of food production is reduced directly through plant diseases. The citrus plants are widely grown fruits worldwide. Each year, a large amount of waste is produced by citrus manufacturers, who annually destroy 50 percent of the citrus peel due to various plant diseases. The discovery of citrus plant diseases and their severity level will improve the quality of agricultural production. Image processing techniques are widely for detection of citrus plant disease. In this paper, evolutionary algorithms are introduced to detect citrus plant diseases and their severity level. The leaf images of citrus plants are collected and those images are pre-processed by removing noise using filtering technique. Then, the diseased portion in the leaf image is extracted by partitioning the image into multiple segments. From the segmented images, the features such as contrast, color, energy, local homogeneity, cluster shade and prominence are extracted using co-occurrence method and these features are processed in GA and PSO to detect the citrus plant diseases and their severity level. In GA, each gene randomly selects the features and classification rule is formed by chromosome. In PSO, each particle randomly selects the features and classification rule is generated. The best classification rule is selected based on the classification accuracy of selected classification rule. After the selection of best classification rule, it is applied to the testing data to detect citrus plant diseases and their severity level.

Why it matches plant phenotyping methods柑橘葉画像から病害の種類と重症度を推定する画像解析手法の開発が中心であり、植物の病害状態を直接評価している。

abstractIn this paper, evolutionary algorithms are introduced to detect citrus plant diseases and their severity level.

Comparative Genomics Screen Identifies Microbe-Associated Molecular Patterns from ' Candidatus Liberibacter' spp. That Elicit Immune Responses in Plants. →

Citrus huanglongbing (HLB), caused by phloem-limited ' Candidatus Liberibacter' bacteria, is a destructive disease threatening the worldwide citrus industry. The mechanisms of pathogenesis are poorly understood and no efficient strategy is available to control HLB. Here, we used a comparative genomics screen to identify candidate microbe-associated molecular patterns (MAMPs) from ' Ca. Liberibacter' spp. We identified the core genome from multiple ' Ca. Liberibacter' pathogens, and searched for core genes with signatures of positive selection. We hypothesized that genes encoding putative MAMPs would evolve to reduce recognition by the plant immune system, while retaining their essential functions. To efficiently screen candidate MAMP peptides, we established a high-throughput microtiter plate-based screening assay, particularly for citrus, that measured reactive oxygen species (ROS) production, which is a common immune response in plants. We found that two peptides could elicit ROS production in Arabidopsis and Nicotiana benthamiana . One of these peptides elicited ROS production and defense gene expression in HLB-tolerant citrus genotypes, and induced MAMP-triggered immunity against the bacterial pathogen Pseudomonas syringae . Our findings identify MAMPs that boost immunity in citrus and could help prevent or reduce HLB infection.

Why it matches plant phenotyping methods植物免疫応答（ROS産生）を定量する柑橘向け高スループット測定法を確立し、候補MAMPのスクリーニングに中心的に用いているため、植物生理状態のフェノタイピング手法として採用。

abstractTo efficiently screen candidate MAMP peptides, we established a high-throughput microtiter plate-based screening assay, particularly for citrus, that measured reactive oxygen species (ROS) production, which is a common immune response in plants.

In-Field Estimation of Orange Number and Size by 3D Laser Scanning →

The estimation of fruit load of an orchard prior to harvest is useful for planning harvest logistics and trading decisions. The manual fruit counting and the determination of the harvesting capacity of the field results are expensive and time-consuming. The automatic counting of fruits and their geometry characterization with 3D LiDAR models can be an interesting alternative. Field research has been conducted in the province of Cordoba (Southern Spain) on 24 ‘Salustiana’ variety orange trees—Citrus sinensis (L.) Osbeck—(12 were pruned and 12 unpruned). Harvest size and the number of each fruit were registered. Likewise, the unitary weight of the fruits and their diameter were determined (N = 160). The orange trees were also modelled with 3D LiDAR with colour capture for their subsequent segmentation and fruit detection by using a K-means algorithm. In the case of pruned trees, a significant regression was obtained between the real and modelled fruit number (R2 = 0.63, p = 0.01). The opposite case occurred in the unpruned ones (p = 0.18) due to a leaf occlusion problem. The mean diameters proportioned by the algorithm (72.15 ± 22.62 mm) did not present significant differences (p = 0.35) with the ones measured on fruits (72.68 ± 5.728 mm). Even though the use of 3D LiDAR scans is time-consuming, the harvest size estimation obtained in this research is very accurate.

Why it matches plant phenotyping methods3D LiDARとK-means分割により、樹上果実数・直径という植物器官形質を推定し、実測値との回帰・比較で技術検証しているため、フェノタイピング手法が中心である。

abstractThe automatic counting of fruits and their geometry characterization with 3D LiDAR models can be an interesting alternative.

Setting up a methodology to distinguish between green oranges and leaves using hyperspectral imaging →

The estimation of green citrus fruit yield is a key parameter for growers and the industry. The early estimation of orange yield at the immature green stage could influence the future market price and allow producers to plan the harvest in advance, thus reducing costs. This research can be considered as a preliminary step for designing low-cost spectral cameras capable of being mounted on unmanned aerial vehicles (UAVs) to estimate orange yield and defects. Images were acquired from oranges and leaves from an orchard in Jeju Island (Jeju, Republic of Korea), using two hyperspectral reflectance imaging systems, one working in the range 400–1000 nm (visible/near infrared, Vis/NIR) and the other between 900 and 2500 nm (short-wave infrared, SWIR). The main objective of the research was to set up a methodology to select the relevant bands - from the two spectral ranges studied - to distinguish between green oranges and leaves and to detect defects, which will allow citrus yield to be estimated. Analysis of variance (ANOVA) and principal component analysis (PCA) were used to select the key wavelengths for this purpose; next, a band ratio coupled with a simple thresholding method was applied. This study showed that the Vis/NIR hyperspectral imaging correctly classified 96.97% and 92.93% of the pixels, respectively, to distinguish between green oranges and leaves and to detect defects, while with the SWIR system, the percentage of pixels correctly classified for these two objectives were 74.79% and 89.31%, respectively. These results confirm that it is possible to use a low number of wavelengths to estimate harvest yield in oranges, which could pave the way for the future development of low-cost and low-weight equipment for the detection of green and sound fruit.

Why it matches plant phenotyping methods柑橘果実と葉の識別および欠陥検出を目的に、ハイパースペクトル画像の波長選択・バンド比・閾値処理を開発しており、収量推定に用いる果実表現型取得法が中心である。

abstractThe main objective of the research was to set up a methodology to select the relevant bands - from the two spectral ranges studied - to distinguish between green oranges and leaves and to detect defects

Field detection and classification of citrus Huanglongbing based on hyperspectral reflectance →

Citrus Huanglongbing (HLB), also called citrus greening, is the most destructive disease in the citrus industry. Detecting the disease as early as possible and then eradicating infected roots can effectively control its spread. For the Shatangju mandarin cultivar, a non-destructive citrus HLB field detection method based on hyperspectral reflectance is proposed in this study. A characteristic band extraction method based on entropy distance and sequential backward selection is explored. Several machine learning algorithms (logistic regression, decision tree, support vector machine, k-nearest neighbor, linear discriminant analysis, and ensemble learning) were used to discriminate between disease groups: healthy, symptomatic HLB-infected, and asymptomatic HLB-infected, based on leaf reflectance. The results showed that the use of primary hyperspectral reflectance is very feasible for such classification. The band selection method proposed in this study provides an option for dimensionality reduction while still providing high classification accuracy. In three-group classification, the SVM learner achieved 90.8% accuracy, while in two-group classification (healthy vs symptomatic HLB leaves), the accuracy reached to 96%. The results also show that using only a few bands is insufficient for classification. In this study, 13 characteristic bands extracted by the proposed method provided the best performance.

Why it matches plant phenotyping methods葉のハイパースペクトル反射から柑橘HLB感染状態を非破壊推定し、特徴波長抽出と分類性能を評価する手法が研究の中心であるため、植物病害フェノタイピング手法として含める。

abstracta non-destructive citrus HLB field detection method based on hyperspectral reflectance is proposed in this study.

Gas Biosensor Arrays Based on Single-Stranded DNA-Functionalized Single-Walled Carbon Nanotubes for the Detection of Volatile Organic Compound Biomarkers Released by Huanglongbing Disease-Infected Citrus Trees. →

Volatile organic compounds (VOCs) released by plants are closely associated with plant metabolism and can serve as biomarkers for disease diagnosis. Huanglongbing (HLB), also known as citrus greening or yellow shoot disease, is a lethal threat to the multi-billion-dollar citrus industry. Early detection of HLB is vital for removal of susceptible citrus trees and containment of the disease. Gas sensors are applied to monitor the air quality or toxic gases owing to their low-cost fabrication, smooth operation, and possible miniaturization. Here, we report on the development, characterization, and application of electrical biosensor arrays based on single-walled carbon nanotubes (SWNTs) decorated with single-stranded DNA (ssDNA) for the detection of four VOCs-ethylhexanol, linalool, tetradecene, and phenylacetaldehyde-that serve as secondary biomarkers for detection of infected citrus trees during the asymptomatic stage. SWNTs were noncovalently functionalized with ssDNA using π-π interaction between the nucleotide and sidewall of SWNTs. The resulting ssDNA-SWNT hybrid structure and device properties were investigated using Raman spectroscopy, ultraviolet (UV) spectroscopy, and electrical measurements. To monitor changes in the four VOCs, gas biosensor arrays consisting of bare SWNTs before and after being decorated with different ssDNA were employed to determine the different concentrations of the four VOCs. The data was processed using principal component analysis (PCA) and neural net fitting (NNF).

Why it matches plant phenotyping methodsHLB感染柑橘が放出するVOCを用いて、無症状感染樹の病態を検出するガスバイオセンサーアレイを開発・特性評価・適用しており、植物状態の取得方法が中心である。

abstractHere, we report on the development, characterization, and application of electrical biosensor arrays based on single-walled carbon nanotubes (SWNTs) decorated with single-stranded DNA (ssDNA) for the detection of four VOCs

In-field citrus detection and localisation based on RGB-D image analysis →

In-field citrus detection and localisation are highly challenging tasks due to varying illumination conditions, partial occlusion of citrus, and the colour variation of citrus at different stages of maturity. A reliable algorithm based on red-green-blue-depth (RGB-D) images was developed to detect and locate citrus in real, outdoor orchard environments for robotic harvesting. A depth filter and a Bayes-classifier-based image segmentation method were first developed to exclude as many backgrounds as possible. A density clustering method was then used to group adjacent points in the filtered RGB-D images into clusters, where each cluster represents a possible citrus. A colour, gradient, and geometry feature-based support vector machine classifier was trained to remove false positives. To test the method, a dataset with 506 RGB-D images was acquired in a citrus orchard on sunny and cloudy days. Results showed that the proposed algorithm was robust with an F1 score of 0.9197; the positioning errors in the x, y and z directions were 7.0 ± 2.5 mm, −4.0 ± 3.0 mm and 13.0 ± 3.0 mm, respectively, and the sizing error was −1.0 ± 4.0 mm. These excellent performance values demonstrate that the proposed method could be used to guide a citrus-harvesting robot.

Why it matches plant phenotyping methodsRGB-D画像から柑橘果実を検出・位置決めし、さらに果実サイズを推定する手法を開発・検証しており、収穫対象の単なる位置検出を超えた器官形質測定が中心です。

abstractA reliable algorithm based on red-green-blue-depth (RGB-D) images was developed to detect and locate citrus in real, outdoor orchard environments for robotic harvesting.

Freeze-Damage Detection in Lemons Using Electrochemical Impedance Spectroscopy. →