Plant phenotyping methods.

Video-based fruit detection and tracking: effects of scanning conditions on fruit load estimation →

Automated fruit counting and yield estimation systems are necessary for efficient orchard management. This study presents a computer vision system based on video multi-object-tracking for fruit load estimation in apple orchards and provides a comprehensive analysis of the system performance under diverse scanning conditions. The system integrates fruit detection, tracking, localization within orchard, and fruit load map generation. Experiments were carried out in an experimental apple orchard containing 420 apple trees. Data was collected with two different RGB-D sensors (Azure Kinect DK and ZED 2) at three different scanning distances (125 cm, 175 cm, and 225 cm) on two different dates prior to the harvest. Comparing the two evaluated sensors, Azure Kinect provided more consistent performance across different dates. Results also show that the longer scanning distance improves accuracy due to seeing the full tree view gives better fruit counts than close partial views. Between the two dates, best results were achieved near harvest due to fruit color at this stage, achieving a Mean Absolute Percentage Error (MAPE) of 6.91 % and a determination coefficient (R 2 ) of 0.733 (using ZED2 sensor at 225 cm distance). Finally, a test comparing scanning from one or both sides of the tree row showed that bilateral scanning improved fruit load estimation at the stretch level by incorporating information from both sides of the canopy. The results of this work demonstrate the effectiveness of the video fruit tracking systems as a useful tool for automating fruit load estimation.

Why it matches plant phenotyping methods動画ベースの果実検出・追跡手法を開発・評価し、リンゴ樹の果実負荷量を推定することが研究の中心であるため、植物フェノタイピング手法として含める。

abstractThis study presents a computer vision system based on video multi-object-tracking for fruit load estimation in apple orchards and provides a comprehensive analysis of the system performance under diverse scanning conditions.

A two-dimensional axis estimation method for pickable canopy apples based on YOLO cascade network. →

Introduction In a complex orchard environment, canopy apples are obscured by various factors, making it hard for apple harvesting robots to accurately determine which apples can be directly harvested. Furthermore, the complex obstruction leads to difficulties in identifying keypoints on the apples and caculating the axis direction, directly affecting the robot's determination of grasping positions. Methods To solve these issues, a two-dimensional (2D) axis estimation method for pickable canopy apples based on a YOLO cascade network was proposed. Firstly, the study introduced SPDConv for lossless downsampling and adopts dynamic upsampling to improve segmentation boundary accuracy, constructing an instance segmentation network named the YOLO-SD model to select pickable apples according to different occlusion conditions and growth states. Secondly, the geometric center of the mask image was located using its minimum enclosing circle, and the region of interest was extracted through morphological dilation. Then, by integrating the RFAConv, SCSA attention, and MBConv modules, a keypoint detection network YOLO-RSM was constructed to extract keypoints of pickable apples. Finally, a 2D axis construction strategy was proposed, which adaptively constructs the growth axis based on the visibility of keypoints. Results Experimental results show that the overall average accuracy mAP50 of the YOLO-SD model for apple segmentation reached 95.2%, and the parameter quantity was reduced to 2.47 M. The average accuracy of the YOLO-RSM model for keypoint detection has reached 90.3%, which is 2.4%, 2.6%, and 4.7% higher than that of the YOLOv8n, YOLO11n, and YOLO12n models respectively. The 2D axis estimation algorithm has an average axis angular error of 7.23° ± 16.73°, and an axis estimation accuracy of 92.68%. Discussion The proposed method can achieve high-precision canopy apple segmentation, keypoint detection, and 2D axis estimation, thus offering technical support for the picking operations of apple harvesting robots.

Why it matches plant phenotyping methodsリンゴ果実のセグメンテーション、キーポイント抽出、成長軸（器官形態）の推定を中心に新規画像解析法を開発・検証しており、単なる収穫対象の検出を超える植物器官形質の推定に該当する。

abstracta two-dimensional (2D) axis estimation method for pickable canopy apples based on a YOLO cascade network was proposed.

In-situ visualisation of the micromechanical deformation of apple tissue using 4D X-ray computed tomography with digital volume correlation. →

Continuous X-ray computed tomography (XCT) combined with digital volume correlation (DVC) is presented to quantify internal three-dimensional strain and failure dynamics in apple cortex tissue during compression, revealing how the cellular microstructure governs its mechanical response. We introduce a dimensionless number Mi that is a function of the average interfacial contact area of cells, cell wall thickness, the average cell volume and tissue porosity, to describe tissue microstructure over different development stages. Mechanical softening during maturation aligned strongly with decreasing Mi, linking microstructure to effective Young's modulus, peak stress, and toughness. DVC revealed distinctive strain-distribution signatures: in young, low-porosity tissue, strain was initially diffuse with early-onset localization indicating progressive failure, whereas mature, high-porosity tissue exhibited sharply peaked strain distributions and highly localized fracture planes indicative of brittle collapse. These findings demonstrate how pore evolution, anisotropy, and cell packing jointly determine macroscopic deformation, establishing XCT-DVC as a powerful framework for connecting plant tissue architecture to mechanical function.

Why it matches plant phenotyping methods4D XCTとDVCを用いてリンゴ組織の内部三次元ひずみ、微細構造、破壊状態を定量化する手法が研究の中心であり、植物組織の構造・力学的状態を抽出しているため。

abstractContinuous X-ray computed tomography (XCT) combined with digital volume correlation (DVC) is presented to quantify internal three-dimensional strain and failure dynamics in apple cortex tissue during compression

Deep Learning Techniques for Crop Health Monitoring and Disease Detection →

This paper explores how deep learning methods can be used to monitor the health of crops and identify diseases, particularly for the apple crop. As the need for food security and sustainable farming methods increases, there is a strong demand for early detection of crop diseases. We have used a Convolutional Neural Network (CNN), based on the model of VGG16 architecture, since the model is known to be effective in image classification. The dataset contained 7771 training images for to enhance machines deep learning regarding plant diseases. Following that, validation image collection of 1747 images divided into four health conditions of the apple crops. In addition, the model was evaluated using 196 new images as a final test. To enhance the model capacity for recognizing diseases of real leaves rather than just memorize exact training pictures, data augmentation was used with ImageDataGenerator of TensorFlow. This means the training images were zoomed, rotated, and shifted to enable the machine detects more variations. Ten epochs of training were performed to measure the model accuracy. The results indicated that the model obtained significant improvement in training and validation accuracy from 56.43% to 78.12% and 92.94 to 96.93%, respectively. Most impressively, the final test dataset, which contained completely new images, scored an accuracy rate of 98%. The results indicate that in the architecture field the application of deep learning methodologies is effective, suggesting that automated detection of diseases by using sophisticated image analysis manages crop diseases identification efficiently. Combination of these methods successfully creates avenues for novel research to built real-time systems of crop disease monitoring, which help farmers increase their productions and farm their lands sustainably.

Why it matches plant phenotyping methodsリンゴ葉の画像から病害・健全状態をCNNで推定する画像ベース植物フェノタイピング手法であり、学習・検証・未知画像で性能評価を行っているため含める。

abstractWe have used a Convolutional Neural Network (CNN), based on the model of VGG16 architecture, since the model is known to be effective in image classification.

APLICAÇÃO DE REDES NEURAIS CONVOLUCIONAIS PARA DETECÇÃO DE DOENÇAS EM FOLHAS DE MACIEIRAS →

O cultivo de maçãs tem grande importância econômica no setor agropecuário brasileiro, especialmente na região Sul do país. No entanto, a produtividade dos pomares é frequentemente comprometida por doenças foliares que, se não tratadas, podem resultar em perdas substanciais. Nesse contexto, os avanços em técnicas de Aprendizado de Máquina têm possibilitado o desenvolvimento de soluções computacionais que auxiliam no diagnóstico agropecuário com maior precisão e agilidade. Este trabalho propõe uma abordagem baseada em Redes Neurais Convolucionais que utiliza aprendizado por transferência para detectar automaticamente sintomas de doenças em folhas de macieira. A metodologia desenvolvida inclui a segmentação e análise de regiões sintomáticas para reduzir o ruído proveniente de áreas saudáveis ​​e direcionar o aprendizado do modelo para sinais relevantes. Um total de 32.382 manchas de sintomas foram extraídas de 1.995 imagens originais, abrangendo cinco classes de distúrbios foliares: glomerela, sarna, danos por herbicidas, deficiência de magnésio e deficiência de potássio. A rede MobileNetV2, treinada por meio de aprendizado por transferência, alcançou um F1-score de 0,926 e 93,8% de acurácia no conjunto de teste reservado. Os resultados indicam um bom desempenho no contexto avaliado, sugerindo o potencial da abordagem como ferramenta de apoio ao diagnóstico da saúde das plantas e à tomada de decisões em campo. Palavras-chave: aprendizado de máquina; visão computacional; doenças em plantas. Application of convolutional neural networks for disease detection in apple tree leaves ABSTRACT: Apple cultivation holds significant economic importance in the Brazilian agricultural sector, especially in the southern region of the country. However, orchard productivity is often compromised by foliar diseases, which, if left untreated, can lead to substantial losses. In this context, advances in Machine Learning techniques have enabled the development of computational solutions that support agricultural diagnostics with greater accuracy and agility. This work proposes an approach based on Convolutional Neural Networks that uses transfer learning to automatically detect disease symptoms in apple leaves. The developed methodology includes segmentation and analysis of symptomatic regions to reduce noise from healthy areas and focus the model’s learning on relevant signals. A total of 32,382 symptom patches were extracted from 1,995 original images, covering five foliar disorder classes: glomerella, scab, herbicide damage, magnesium deficiency, and potassium deficiency. The MobileNetV2, trained via transfer learning, achieved a F1-score of 0.926 and 93.8% accuracy on the held-out test set. The results indicate good performance in the evaluated setting, suggesting the approach’s potential as a tool to support plant-health diagnosis and field decision-making. Keywords: machine learning; computer vision; plant disease.

Why it matches plant phenotyping methods葉の病徴を画像から自動検出・分類するCNNと、症状領域のセグメンテーションを中心的に開発・評価しており、植物病害状態の画像ベース表現型計測に該当する。

abstractEste trabalho propõe uma abordagem baseada em Redes Neurais Convolucionais que utiliza aprendizado por transferência para detectar automaticamente sintomas de doenças em folhas de macieira.

A Comparative Study of Convolutional Neural Network based Transfer Learning Models for Plant Disease Detection →

Background: Plant diseases significantly threaten global food security, reducing potential harvests and sometimes causing total crop failure. Traditional detection methods, which rely on manual inspection and laboratory testing, are time-consuming, costly and prone to human error. Methods: To address these challenges, this study applies transfer learning techniques using deep convolutional neural networks for accurate and efficient plant disease detection. A comparative analysis of nine pretrained models, VGG16, VGG19, ResNet50, ResNet101V2, MobileNetV2, InceptionV3, DenseNet121, InceptionResNetV2 and Xception was conducted on the PlantVillage dataset, focusing on apple, potato and peach leaf images. Result: Results show that DenseNet121 and ResNet101V2 achieved the highest accuracy, particularly for potato leaves with 98.5%, while MobileNetV2 also performed well with up to 99% accuracy for apple and peach leaves. The study demonstrates that transfer learning effectively enhances plant disease classification, enabling faster, more reliable and resource efficient detection for precision agriculture.

Why it matches plant phenotyping methods植物葉画像から病害状態を推定する深層学習手法を比較評価しており、病害表現型の取得・分類が研究の中心である。

abstractthis study applies transfer learning techniques using deep convolutional neural networks for accurate and efficient plant disease detection.

Apple Tree Distance and Volume Measurement Using LiDAR and RGB-D Imaging →

LiDAR (Light Detection and Ranging) and RGB-D camera imaging have emerged as essential tools in agricultural applications, particularly for plant size and distance measurements, enabling non-destructive, cost-effective, and precise estimation. The objective of this study was to measure the plant canopy dimensions and distance between apples using commercial LiDAR, and an RGB-D camera with a speed sprayer platform was used to determine whether LiDAR provides a higher measurement accuracy under field conditions. Data were collected in an apple orchard in Muju, Republic of Korea. Commercial 3D LiDAR, a terminal box, an RGB-D camera, a microcontroller, a power supply, and individual display monitors were integrated into a customized data acquisition (DAQ) box for LiDAR point cloud (PCD), RGB, and depth imagery data collection. Commercial software was used for data acquisition, data conversion (pcap to PCD), segmentation of regions of interest (ROI), and pre-processing of data. PCD processing and measurement consisted of data frame selection, data conversion, outlier removal, downsampling, denoising, ground point removal by filtering, voxelization, and density map generation using an open access programming language script. Depth image processing included importing raw data, shaping metadata using intrinsic camera parameters, visualizing depth images, extracting depth points, and measuring the plant canopy at the pixel level. RGB image analysis involved grayscale conversion, thresholding, segmentation of ROI, contour preparation, noise removal, and binary masking for eliminating the background. Estimated results were compared to measured results. LiDAR measurements showed the closest agreement with the measured results for plant height, canopy volume, plant spacing, and row distance, outperforming both RGB and depth imaging. Under field conditions, plant spacing and row distance were estimated with accuracies of 97.5% and 94.7%, respectively, exhibiting higher measurement accuracies than RGB and depth imagery data results. Despite some discrepancies due to complex plant geometry and dynamic data collection, the results support data collection strategies critical for precision horticulture.

Why it matches plant phenotyping methodsLiDARとRGB-D画像を用いたリンゴ樹の樹冠寸法・体積・樹間距離・列間距離の取得と精度比較が研究の中心であり、植物表現型計測手法の開発・検証に該当する。

abstractThe objective of this study was to measure the plant canopy dimensions and distance between apples using commercial LiDAR, and an RGB-D camera with a speed sprayer platform was used to determine whether LiDAR provides a higher measurement accuracy under field conditions.

ALSDet: a global context-enhanced network for detecting small-target diseases on apple leaves. →

Accurate detection of small-target diseases on apple leaves is of great importance for optimizing orchard management and facilitating precision agriculture. Small-target disease detection remains challenging due to insufficient features and complex backgrounds. This work proposes an effective detector for apple leaf small-target diseases called ALSDet. The global context module is integrated into Stage2 to Stage4 of the ResNet-50 backbone, yielding a refinement of the feature-extraction architecture. In the bottleneck blocks of the Stage2 and Stage3, dilated convolution is used in place of the normal 3×3 convolution to enlarge the receptive field for small targets and strengthen feature extraction. During the model training phase, a multi-scale training strategy combined with the online hard example mining method is adopted to focus on learning hard samples and enhance adaptability for various scale targets. According to the experimental results, ALSDet obtains a mean average precision (mAP) of 65.6% and an average recall (AR) of 71.2% on the dataset. The proposed model achieves the highest levels in both mAP and AR when compared to the popular object detection models, such as Cascade R-CNN, Faster R-CNN, GFL, Grid R-CNN, Libra R-CNN, FCOS, VFNet, RetinaNet, SSD, YOLOv7, and YOLOv8. For small-target diseases like rust and frog eye leaf spot, the average precision surpasses 87% with an intersection over union (IoU) threshold of 0.5. These results confirm that ALSDet achieves stable performance against existing methods, demonstrating its potential as a practical tool for intelligent orchard disease management.

Why it matches plant phenotyping methodsリンゴ葉の病斑を画像から検出する深層学習モデルを開発・比較評価しており、植物の病害状態の取得・推定が研究の中心である。

abstractThis work proposes an effective detector for apple leaf small-target diseases called ALSDet.

Deep Learning Techniques for Crop Health Monitoring and Disease Detection →

This paper explores how deep learning methods can be used to monitor the health of crops and identify diseases, particularly for the apple crop. As the need for food security and sustainable farming methods increases, there is a strong demand for early detection of crop diseases. We have used a Convolutional Neural Network (CNN), based on the model of VGG16 architecture, since the model is known to be effective in image classification. The dataset contained 7771 training images for to enhance machines deep learning regarding plant diseases. Following that, validation image collection of 1747 images divided into four health conditions of the apple crops. In addition, the model was evaluated using 196 new images as a final test. To enhance the model capacity for recognizing diseases of real leaves rather than just memorize exact training pictures, data augmentation was used with ImageDataGenerator of TensorFlow. This means the training images were zoomed, rotated, and shifted to enable the machine detects more variations. Ten epochs of training were performed to measure the model accuracy. The results indicated that the model obtained significant improvement in training and validation accuracy from 56.43% to 78.12% and 92.94 to 96.93%, respectively. Most impressively, the final test dataset, which contained completely new images, scored an accuracy rate of 98%. The results indicate that in the architecture field the application of deep learning methodologies is effective, suggesting that automated detection of diseases by using sophisticated image analysis manages crop diseases identification efficiently. Combination of these methods successfully creates avenues for novel research to built real-time systems of crop disease monitoring, which help farmers increase their productions and farm their lands sustainably.

Why it matches plant phenotyping methodsリンゴ葉画像から健康状態・病害を推定するCNN画像解析手法が研究の中心であり、学習・検証・新規画像での評価も実施しているため、植物病害表現型の手法研究として含める。

abstractThis paper explores how deep learning methods can be used to monitor the health of crops and identify diseases, particularly for the apple crop.

Detect Plant Disease and Recommend Fertilizer and Supplement Using CNN & Mobile Net Algorithm →

Agriculture plays a crucial role in the Indian economy. Early detection of plant diseases is very much essential to prevent crop loss and further spread of diseases. Most plants such as apple, tomato, cherry, grapes show visible symptoms of the disease on the leaf. These visible patterns can be identified to correctly predict the disease and take early actions to prevent it. This can be overcome by the use of machine learning and deep learning algorithms. Hence, we are proposing a method that which is detecting the disease of a tomato plant from their leaf images. Here the process is performed with the deep learning algorithms Convolutional Neural Network (CNN), and MobileNet which is a one of the transfer learning method of CNN. Once after training the dataset with the algorithms, the accuracy of algorithms is compared and the images are classified. And the precautions are also provided for the classified plant.

Why it matches plant phenotyping methodsトマト葉画像から病害状態をCNN/MobileNetで推定・分類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。

abstractwe are proposing a method that which is detecting the disease of a tomato plant from their leaf images

Comprehensive plant disease classification and severity estimation for sustainable farming via automatic segmentation and multi-scale feature fusion →

Abstract Detecting plant leaf diseases at an early stage is one of the most important requirements for sustainable agriculture, increasing crop productivity, and achieving the global Sustainable Development Goals (SDGs). However, accurately recognizing them in real-world farm fields can still be difficult due to factors such as background complexity, changes in light conditions, and very similar looking classes from a visual standpoint. In order to solve these problems, the authors here present a new Multi-Scale Feature Fusion (MSFF) model that can offer robust and highly accurate performance in identifying plant leaf diseases. Firstly, the brand new hybrid method starts with a U-Net segmentation designed exclusively to separate the diseased parts and thus allow the classification to be more robust. Next, Rank Order Fuzzy (ROF) is implemented to get rid of the background while still maintaining the edges, and the additional data is used for the network to generalize better. The color distribution is then analyzed to determine the variations in color brought about by the infection. In terms of features, EfficientNet is paired with an Attention-based Autoencoder to produce both spatially detailed global features and compact latent representations. The two sets of features are then combined through Canonical Correlation Analysis (CCA) which not only identifies the dependencies between the features but also enhances the discriminative strength. The final fused feature set is fed to module based on YOLO for detection and classification in order to obtain the final result of the plant disease identification system which is both accurate and fast. Among other datasets, the model has been tested on different apple leaf datasets such as FGVC7, AppleLeafSet, PlantVillage Apple, Kaggle Apple Leaves, and ATLDSD, which contain five disease classes. The results of the experiments indicate a classification accuracy of 99.95%, thus the model is superior to several state-of-the-art deep learning and classical machine learning methods. Statistical methods like five-fold cross-validation, paired t-tests ( p

Why it matches plant phenotyping methods植物葉の病変部位を画像から自動抽出し、病害の識別・分類を行う画像ベースの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。

abstractthe authors here present a new Multi-Scale Feature Fusion (MSFF) model that can offer robust and highly accurate performance in identifying plant leaf diseases

Apple Crop Health Detection Based on Vegetation Indices at Shalimar, Kashmir: North-Western Himalayas →

Apple orchard health monitoring is important for supporting timely crop management under the temperate conditions of Kashmir Valley. The present study evaluated the seasonal behaviour of four Sentinel-2-derived vegetation indices, namely the Normalized Difference Vegetation Index, Soil Adjusted Vegetation Index, Modified Soil Adjusted Vegetation Index and Enhanced Vegetation Index, in a 2 ha apple orchard at SKUAST-K, Shalimar, Kashmir, during the 2020 growing season. Sentinel-2 imagery acquired from April to September was processed using SNAP and QGIS, and mean vegetation index values were extracted for the orchard area. Monthly weather data, including average temperature and rainfall, were obtained from the on-campus meteorological observatory. Ground observations at 20 georeferenced points were used to support the interpretation of canopy development, phenological stage and visible plant health condition. All four vegetation indices showed a seasonal increase from April to July-August, followed by a decline during September-October, corresponding to canopy development, fruit maturation, harvest and senescence. Based on the monthly values presented in the study, temperature showed positive associations with the vegetation indices, with the strongest relationship observed for MSAVI, followed by NDVI and SAVI. Rainfall showed weak and non-significant associations with the indices during the study period. The results indicate that Sentinel-2-derived vegetation indices can reflect seasonal canopy dynamics in apple orchards under the studied conditions. MSAVI appeared particularly useful for representing canopy development, while field observations remained necessary for interpreting pest, disease and phenological effects.

Why it matches plant phenotyping methodsSentinel-2画像から植生指数を算出し、リンゴ樹冠の季節動態・健康状態を評価する測定ワークフローが研究の中心であり、植物状態の推定に直接用いられている。

abstractThe present study evaluated the seasonal behaviour of four Sentinel-2-derived vegetation indices

Double Transfer Learning-Based Capsule Network for Multi-Crop Plant Disease Classification Using Heterogeneous Leaf Image Datasets →

Abstract Crop disease is a major worldwide problem in agricultural production and food security, adversely affecting yield and quality for a variety of plant species. To overcome these drawbacks, this research provides a Double Transfer Learning-based Capsule Network (DTL-CapsNet) approach for automated plant disease classification with multiple crops. Based on the image pre-processing, segmentation, double transfer learning, and Capsule Networks technologies, the proposed framework extracts discriminative features of the diseases and maintains spatial relations between the leaves symptoms effectively. This double transfer learning approach involves extracting general visual features from pre-trained deep learning models and then fine-tuning these features to classify plant diseases. Capsule Networks then leverage the visual similarity of disease patterns to make the recognition more robust, while simultaneously adding hierarchical part–whole relationships in leaf structures, thereby improving the feature representation. Experiments were performed on a heterogeneous data set consisting of 21,927 leaf images belonging to 17 different healthy and diseased classes of apple, chilli, cotton, corn and potato crops. The experimental results proposed DTL-CapsNet framework is more accurate compared to the traditional CNN-based models and conventional transfer learning models. The proposed method of double transfer learning and Capsule Networks offers an efficient and scalable approach for intelligent plant disease diagnosis, offering significant potential in precision agriculture and real-time crop monitoring systems.

Why it matches plant phenotyping methods葉画像から植物病害状態を分類する画像・計算手法が研究の中心であり、提案手法の開発と既存モデルとの比較検証が行われているため。

abstractBased on the image pre-processing, segmentation, double transfer learning, and Capsule Networks technologies, the proposed framework extracts discriminative features of the diseases and maintains spatial relations between the leaves symptoms effectively.

Dual-view weakly-supervised learning for apple tree flower counting. →

This paper presents a method to estimate apple tree flower cluster count using image analysis techniques. The main research question is how accurately flower clusters on apple trees can be counted using a camera-based approach and weakly-supervised learning, compared to traditional visual estimation. A camera is used to capture images of blooming apple trees from two sides. These images are processed by a weakly-supervised model based on a ResNet feature extractor and a feature pyramid network serving as the feature aggregator. The model is trained and validated using reference data obtained through manual flower cluster counts in the orchard and from estimations based on camera images. The model was trained using field-validated data and visually-estimated data, enabling a comparative evaluation. The proposed model surpassed conventional image segmentation methods in estimating both manually counted and visually estimated flower clusters. The proposed method achieves a relative error of 10.26% in estimating flower cluster quantities, demonstrating its effectiveness and improved accuracy over traditional approaches. Its reliance on field-validated reference data adds to its robustness and practical relevance.

Why it matches plant phenotyping methodsリンゴ樹の花房数という植物形質を、カメラ画像と弱教師あり学習で推定する手法の開発・検証が研究の中心であるため。

abstractThis paper presents a method to estimate apple tree flower cluster count using image analysis techniques.

HSGAN-based near-infrared hyperspectral reconstruction from characteristic wavelengths images for apple bruise detection. →

Hyperspectral images contain richer spectral and spatial information than multispectral images, yet traditional equipment suffers from limitations such as bulky size and complex data processing. This study constructs a task-oriented near-infrared (NIR) hyperspectral reconstruction and detection framework for the precise detection of early apple bruises. First, apple samples were collected using a hyperspectral imaging system. The Weight Extremum Method was employed to screen seven characteristic wavelengths (976.4 nm, 1064.8 nm, 1175.8 nm, 1192.5 nm, 1295.4 nm, 1449 nm, and 1631.6 nm), which were further reduced to three key wavelengths (1064.8 nm, 1175.8 nm, and 1449 nm). Based on the datasets constructed from these bands, the HSGAN framework was used as the reconstruction backbone to reconstruct hyperspectral images ranging from 866 nm to 1701 nm. Results demonstrated that reconstruction performance was optimal with seven input bands (PSNR = 37.81, SSIM = 0.973) and remained favorable with three bands (PSNR = 34.70, SSIM = 0.950). Finally, YOLOv11n was used to detect bruises on both original and reconstructed images. Detection accuracy using reconstructed spectra from the 7-band input approached that of the original images (mAP50 = 0.994), while the 3-band input also maintained high precision (mAP50 = 0.992, Recall = 0.993). These results demonstrate that reconstructing 254 NIR bands from just three characteristic wavelengths is feasible. This framework significantly reduces data acquisition costs while enabling high-precision early bruise detection, offering a practical solution for agricultural quality control.

Why it matches plant phenotyping methodsリンゴ果実の打撲（植物器官の状態）を対象に、少数波長画像からNIRハイパースペクトル画像を再構成し、検出性能を評価する画像・計算フェノタイピング手法が研究の中心である。

abstractThis study constructs a task-oriented near-infrared (NIR) hyperspectral reconstruction and detection framework for the precise detection of early apple bruises.

Early detection and firmness prediction of apple fruit infested by Bactrocera dorsalis using hyperspectral imaging combined with 1D convolutional neural network →

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

Why it matches plant phenotyping methodsリンゴ果実の硬度という植物器官形質と食害状態を、ハイパースペクトル画像および1D CNNで推定する手法が題名上の中心である。

titleEarly detection and firmness prediction of apple fruit infested by Bactrocera dorsalis using hyperspectral imaging combined with 1D convolutional neural network

Apple Leaf Disease Detection Based on Improved YOLOv11 with DSSA Mechanism. →

Visual inspection of apple leaf diseases is inefficient and subjective, limiting large-scale orchard applications. To realize rapid and accurate disease identification, this paper proposes an improved YOLOv11 model integrated with a Dual Sparse Selection Attention (DSSA) module. By embedding the DSSA module into the key layers of the YOLOv11 backbone network, the model enhances fine-grained feature extraction for small and complex lesions while suppressing background interference. A tailored training strategy with an optimized learning rate and optimizer is designed to ensure stable convergence. Experiments are conducted on a dataset consisting of 7594 images covering four categories: black rot, rust, scab, and healthy leaves. The proposed model achieves precision of 0.973, recall of 0.978, mAP50 of 0.991, and 0.949 mAP50-95, outperforming YOLOv8, YOLOv9, YOLOv10, and the vanilla YOLOv11. Furthermore, a Qt-based visualization system is developed for practical orchard deployment. This method provides a reliable solution for intelligent apple leaf disease detection and smart orchard management.

Why it matches plant phenotyping methodsリンゴ葉の病斑・健全状態を画像から推定する検出モデルを開発・比較し、実用システムまで構築しており、植物病害表現型の取得手法が中心である。

abstractVisual inspection of apple leaf diseases is inefficient and subjective, limiting large-scale orchard applications.

Research on digital fruit tree reconstruction method based on neural radiance field theory →

The objective of this study is to propose a digital fruit tree reconstruction method based on neural radiation field theory that will enable the efficient, accurate, and non-destructive acquisition of phenotypic information from fruit trees, while simultaneously reducing the cost of collecting this data. Firstly, a low-cost information acquisition platform is constructed for the purpose of shooting a multi-view video around a fruit tree. The video is then extracted and framed using a motion recovery structural algorithm, thereby obtaining a multi-view image sequence of the fruit tree with positional data. Secondly, the the image sequence is employed to train the neural radiation field, thereby obtaining a converged three-dimensional scene of the fruit tree. Ultimately, the three-dimensional scene is derived in the form of a point cloud, thus yielding a high-phenotypic detail point cloud model of the fruit tree. A multi-period point cloud model of an apple tree in an orchard environment, encompassing the flowering, fruiting, and dormant periods, was reconstructed using the aforementioned method. The experimental results demonstrated that the error associated with the tree shape data recorded by the multi-period point cloud model established by the aforementioned method was, on average, below 5%. Furthermore, the average error across all periods was 2.69%, representing a 75.50% reduction compared to traditional reconstruction methods. The dimensional accuracy of the point cloud model at the organ scale can reach the millimetre level, with an average error of 3.10% for fruit diameter, which is 66.19% lower than that of the traditional reconstruction method. The reconstruction method is robust to all periods of fruit trees and can meet the majority of cases of digital fruit tree reconstruction.

Why it matches plant phenotyping methods果樹の多視点画像からNeRFと点群を用いて樹形・器官寸法などの表現型を非破壊取得する手法を開発・検証しており、方法が研究の中心である。

abstractpropose a digital fruit tree reconstruction method based on neural radiation field theory that will enable the efficient, accurate, and non-destructive acquisition of phenotypic information from fruit trees

Apple Origin Classification and Sugar Content Prediction of 'Fuji' Apples Using Near-Infrared Spectroscopy and Deep Learning. →

Accurate apple origin identification and non-destructive internal quality evaluation are important for fruit traceability, quality grading, and post-harvest management. Unlike previous studies mainly focusing on origin classification, this study established a dual-task near-infrared spectroscopy framework integrating geographical origin classification and soluble solid content (SSC, °Brix) prediction for Fuji apples. Samples were collected from three representative production regions in China: Alar in Xinjiang, Yantai in Shandong, and Luochuan in Shaanxi. Near-infrared diffuse reflectance spectra were acquired from 375 apples, generating 3000 spectral samples for origin classification and 750 SSC-calibrated samples for sugar content prediction. For classification, six deep learning models were evaluated using standardized full-spectrum input without chemometric spectral preprocessing, and the Transformer achieved the best performance, with a test accuracy of 96.22%. For SSC regression, spectra were preprocessed using standard normal variate and Savitzky-Golay filtering. The DNN model achieved the best prediction performance, with MAE = 0.5958 °Brix, RMSE = 0.7333 °Brix, R 2 = 0.8646, and Pearson r = 0.9338. These results indicate that near-infrared spectroscopy combined with deep learning can support both Fuji apple origin authentication and non-destructive local tissue SSC assessment.

Why it matches plant phenotyping methodsリンゴ果実のSSC（糖度）という植物器官形質を、近赤外分光と深層学習で非破壊推定する方法を構築・評価しており、表現型取得手法が研究の中心である。

abstractthis study established a dual-task near-infrared spectroscopy framework integrating geographical origin classification and soluble solid content (SSC, °Brix) prediction for Fuji apples.

YOLO-APLD: A Lightweight Apple Leaf Disease Detection Model Based on Multiscale Feature Fusion. →

The precise and timely identification of apple leaf diseases play a key role in targeted pesticide application in orchards. Conventional deep learning techniques encounter issues like the substantial size of model parameters and low detection accuracy across various disease scales in natural environments. To overcome these limitations, this paper presents YOLO-APLD, a lightweight algorithm for detecting apple leaf diseases, utilizing the improved YOLOv8n model. The proposed model incorporates four key improvements to enhance its detection performance. First, an EP-C2f enhancement module is embedded at the output of the backbone to strengthen the representation of local and structural features of damaged area, thereby achieving significant improvements in the recognition of morphologically complex diseases such as rust. Additionally, spatial intersection over union (SIoU) loss and focal loss are combined to form Focal-SIoU loss, which simultaneously optimizes bounding box regression and classification, thus enhancing the detection stability for hard-to-distinguish samples and few-shot categories including mosaic and brown spot. Meanwhile, a bidirectional feature pyramid network is adopted in the neck for efficient multiscale feature fusion, which strengthens the perceptual capability for both large-scale damaged area (powdery mildew and scab) and small-scale damaged area (Alternaria blotch and gray spot). Finally, a Slim-neck structure is employed to simplify the feature fusion architecture, reducing model size and accelerating inference speed. Comprehensive experiments demonstrate that YOLO-APLD achieves excellent performance while maintaining real-time capability, with precision, recall, mean average precision, and F1-score reaching 88.5, 84.3, 88.5, and 86.4%, respectively. Compared with YOLOv8n, these metrics show respective improvements of 1.7, 1.5, 0.8, and 1.6%. Meanwhile, floating point operations, parameter count, and model size are reduced by 22.2, 23.3, and 17.5%, respectively. The detection frame rate on edge computing devices reaches 90.3 f/s, indicating significantly accelerated inference speed. Additionally, testing performance on grape and tomato datasets further validates the generality of the proposed method. In summary, YOLO-APLD exhibits strong detection performance in the field of apple leaf disease detection and can provide practical technical support for precision pesticide application in orchards and on-site disease monitoring.

Why it matches plant phenotyping methodsリンゴ葉の病斑・病害状態を画像から検出するYOLOベース手法を開発・評価しており、植物病害表現型の取得が研究の中心である。

abstractComprehensive experiments demonstrate that YOLO-APLD achieves excellent performance while maintaining real-time capability

Efficient hyperspectral band selection via occlusion-based neural network ranking for detecting fruit-bruise severity →

Hyperspectral imaging (HSI) provides rich spectral information across hundreds of narrow bands, making it a powerful tool for material classification. However, processing all available bands is computationally expensive and often impractical for near real-time applications. In this work, an occlusion-based band-selection method— developed earlier by the authors—is applied to a multiclass classification task to identify the most informative spectral bands for a target task while substantially reducing data dimensionality. For a given application, realistic spectral variations are first simulated through data augmentation under changing intensity and noise conditions. The augmented spectra are then used to train an artificial neural network (ANN) with the full spectral input. Band importance is subsequently evaluated by systematically occluding individual spectral bands and measuring the resulting degradation in classification performance, thereby forming a reduced candidate pool. A computationally manageable exhaustive search is then performed within this pool to identify a smaller subset of bands. As a case study, the method is applied to Honeycrisp apple bruise-severity classification using spectra in the 900–1700 nm range with 336 bands. The full-band ANN achieves 97.7% classification accuracy, while the occlusion-based 16-band and 5-band subsets achieve 89.8% and 80.7%, respectively. Under the same subset sizes, PCA-based selection achieves 84.1% and 74.4%. These results indicate that the proposed method preserves task-relevant spectral information more effectively than the PCA-based baseline, while substantial band reduction can shorten acquisition time, lower computational cost, and support on-device or edge deployment in resource-constrained platforms such as smart cameras.

Why it matches plant phenotyping methodsハイパースペクトル画像からリンゴ果実の bruise severity を推定するためのバンド選択法を中心的に適用・評価しており、植物器官の状態を定量化するフェノタイピング手法に該当する。

abstractan occlusion-based band-selection method— developed earlier by the authors—is applied to a multiclass classification task to identify the most informative spectral bands for a target task while substantially reducing data dimensionality.

Unleashing the power of visible-near infrared spectroscopy: predicting Golden Delicious apple enzyme activity. →

Background Enzymatic browning is a significant reaction in fruits that affects their color, appearance, and quality. The quality of apples, as a perishable product, is mainly influenced by the activity of two browning-related enzymes, polyphenol oxidase (PPO) and peroxidase (POD), during storage. Assessment of these enzymes using conventional methods is often destructive and time-consuming, preventing rapid and non-invasive monitoring of fruit quality. In this study, a visible-near infrared (visible-NIR) spectroscopy approach was developed to predict the enzymatic activity of PPO and POD in intact Golden Delicious apples, aiming to enable rapid, non-destructive evaluation and to identify the most informative spectral regions for industrial applications. Results Both support vector regression (SVR) and decision tree (DT) algorithms achieved high performance when combined with non-linear feature selection algorithms. The best performance, in terms of elapsed time and figure of merits, was achieved by combining particle swarm optimization (PSO) with SVR and DT. However, partial least squares (PLS) models outperformed both SVR-PSO and DT-PSO. Conclusions This study is an advanced proof of concept of the use of visible-NIR spectroscopy - combined with variable selection and machine learning algorithms - for predicting browning-related enzyme activity in apples. The SVR and DT algorithms, coupled with metaheuristic strategies, reached lower performances than PLS, but the success of the variable selection strategy lays the groundwork for developing a miniaturized sensor for assessing apple quality during storage and controlling browning. © 2026 The Author(s). Journal of the Science of Food and Agriculture published by John Wiley & Sons Ltd on behalf of Society of Chemical Industry.

Why it matches plant phenotyping methodsリンゴ果実の酵素活性という植物器官の状態を、可視近赤外分光と機械学習で非破壊推定する手法を開発・比較検証しており、表現型取得法が中心である。

abstracta visible-near infrared (visible-NIR) spectroscopy approach was developed to predict the enzymatic activity of PPO and POD in intact Golden Delicious apples

An Ai-Powered Based Solution for Automated Plant Disease Detection →

Reducing agricultural yield and food security all over the world are major impacts of diseases on crops. Particularly in developing regions this is becoming a severe issue. Early disease identification followed by necessary steps is the way to control further infection. But, identification of crop diseases on the spot can be quite tough because of the scarcity of skilled agronomists who can recognize different plant diseases. A web application based framework is introduced in this paper for real time, automatic recognition of leaf diseases using an AI application. With this framework the plants which have got infected with 38 categories of diseases over 14 plants of apple, corn, tomato, grape, peach, strawberry, citrus, etc can be detected automatically. Size of the image can be anything but here, 160 x 160 pixel leaf image is used as input to the algorithm. It would helps to identify the category of the plant, disease, probability of detection and reason of the infection along with suggesting the appropriate solutions. It is being developed in Python language with the support of TensorFlow, Keras and flask web application for instant response via an interactive web application with Drag and Drop facility. In this, experimental results show good accuracy to classify and recognize the leaf diseases making this system a smart application for farmers, researcher and the expert.

Why it matches plant phenotyping methods植物葉の画像から病害状態を自動推定するAI手法とWebアプリケーションが中心であり、植物の病徴・病害状態を直接評価するため、植物フェノタイピング手法として収録する。

abstractA web application based framework is introduced in this paper for real time, automatic recognition of leaf diseases using an AI application.

Apple Origin Classification and Sugar Content Prediction of ‘Fuji’ Apples Using Near-Infrared Spectroscopy and Deep Learning →

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

Why it matches plant phenotyping methods近赤外分光と深層学習を用いてリンゴ果実の糖含量を予測する手法が題名の中心であり、植物器官の品質形質を推定するため。

titleApple Origin Classification and Sugar Content Prediction of ‘Fuji’ Apples Using Near-Infrared Spectroscopy and Deep Learning

Early Apple Bruise Detection via Discrete Hyperspectral Signatures with SHAP-Guided Feature Selection and a CNN-Transformer Model. →

Accurate detection of early invisible apple bruises is important for post-harvest quality assessment. Although hyperspectral imaging (HSI) provides rich spectral information, its high dimensionality introduces substantial redundancy and weak-signal interference. This study proposes an integrated framework combining waveband optimization and discrete spectral modeling for efficient bruise detection. A Selection-Refined Improved Grey Wolf Optimization (SR-IGWO) algorithm was developed to select 18 bruise-sensitive wavebands from 273 channels (996-2501 nm), achieving a 93.4% reduction in spectral dimensionality. SHAP analysis was further used to interpret the selected bands in relation to biochemical responses associated with bruising. To address the mismatch between conventional CNNs and sparse discrete spectral inputs, a CNN-Transformer hybrid model (DSFormer) was designed using pointwise convolution for band embedding and a Transformer encoder to capture global dependencies. Experimental results across ten independent runs achieved a classification accuracy of 99.11% ± 0.08%, a recall of 96.04% ± 1.08%, and an F1-score of 95.95% ± 0.39% under the tested conditions. Ablation studies suggest that the proposed architecture supports effective detection under sparse spectral conditions. Although validation was limited to a single cultivar and controlled sampling, the proposed framework provides a promising preliminary exploration of reduced hyperspectral data for non-destructive fruit bruise detection.

Why it matches plant phenotyping methodsリンゴ果実の打撲状態を対象に、ハイパースペクトル波長選択とCNN-Transformerによる症状検出手法を開発・検証しており、植物器官の状態取得が研究の中心である。

abstractThis study proposes an integrated framework combining waveband optimization and discrete spectral modeling for efficient bruise detection.

Development of a portable online nondestructive detection device for apple watercore based on visible/near-infrared spectroscopy. →

Visible and near-infrared (Vis/NIR) spectroscopy has been widely applied in fruit quality detection due to its advantages of rapid efficiency, non-invasiveness, and suitability for detecting opaque samples. To address the issue of whether apple watercore occurs during the growth and maturation of apples, a portable on-line nondestructive detection device based on Vis/NIR spectroscopy was designed to achieve accurate detection of apple watercore. The device employs the AIOX2000-13 spectrometer as the detection unit, with an STM32F103VET6 ARM-based processor as the main control chip, and integrates a 4G wireless communication module to establish a stable data transmission channel between the processor and the computer. This structure ensures the efficient and stable transmission of apple spectral data and detection results, thereby meeting the need for in-field nondestructive detection of apple watercore on apple trees. The system is based on a self-designed spectral data acquisition mechanism and uses a transmission detection method to collect spectral data from 500 'Fuji' apple samples in two directions. The spectral data were preprocessed using Standard Normal Variate (SNV), and the dataset was divided using the Spectral Projection based on X-Y distances (SPXY) algorithm. Important feature wavelengths related to apple watercore were extracted by combining the Uninformative Variable Elimination method with the Successive Projections Algorithm (UVE-SPA). Subsequently, a detection model, SNV-UVE-SPA-SVM, was constructed using a Support Vector Machine (SVM) optimized by the Honey Badger Algorithm (HBA), achieving a test set accuracy of 96%. After research and analysis, Direction 1 was identified as the optimal acquisition direction, and field verification was conducted on 50 apple samples, with a detection accuracy of 94%. The results show that the detection device has the advantages of portability, high efficiency, and suitability for in-field detection, making it suitable for the rapid in-field detection of apple watercore.

Why it matches plant phenotyping methodsリンゴの水心症という植物状態を対象に、可視・近赤外分光による携帯型非破壊検出装置と解析モデルを開発し、圃場検証まで実施しており、表現型取得手法が中心である。

abstracta portable on-line nondestructive detection device based on Vis/NIR spectroscopy was designed to achieve accurate detection of apple watercore.

Interpretable hyperspectral analysis of soluble solids content in apples: spectral attribution and mechanistic insights from linear and deep learning models. →

Soluble solids content (SSC) is a key determinant of apple sweetness and market quality, and its rapid, nondestructive assessment is essential for postharvest grading. Hyperspectral imaging (HSI) provides rich spectral information for SSC prediction; however, conventional wavelength selection strategies are largely data-driven and lack interpretability, while the underlying mechanisms of spectral information utilization across different modeling approaches remain insufficiently understood. In this study, an interpretable hyperspectral analysis framework was developed to investigate both effective wavelength selection and model-dependent spectral response mechanisms. Average spectra extracted from the pulp region of interest (ROI) were used for analysis. A linear model (PLSR) combined with SHAP (SHapley additive exPlanations) was employed to quantify global feature contributions, while a one-dimensional convolutional neural network with dual-attention mechanisms (BrixCNN) was constructed and interpreted using integrated gradients (IG) to capture nonlinear spectral dependencies. The consistency and divergence between the two attribution strategies were further quantitatively analyzed. Results showed that both SHAP-PLSR and IG-BrixCNN identified informative wavelength subsets that significantly reduced spectral dimensionality while maintaining comparable predictive performance (R 2 ≈ 0.83-0.84, RPD > 2.4). Despite similar predictive accuracy, the two models exhibited distinct spectral utilization patterns: The linear model primarily relied on dominant, high-variance spectral variations, whereas the deep learning model captured weaker, more distributed, and nonlinear spectral patterns. Meanwhile, partial overlap in the 1100-1300 nm region suggested that both models may utilize correlated spectral variations within similar wavelength domains for SSC prediction. These findings indicate that comparable predictive performance can arise from distinct yet complementary spectral utilization patterns, reflecting model-dependent information extraction mechanisms rather than direct chemical specificity. This study provides new insights into wavelength selection strategies and enhances the interpretability and reliability of hyperspectral analysis for fruit quality assessment.

Why it matches plant phenotyping methodsリンゴ果実のSSCという植物器官形質を対象に、ハイパースペクトル画像、波長選択、SHAP/IG解釈を統合した予測・解析フレームワークを開発しており、形質取得手法が研究の中心である。

abstractIn this study, an interpretable hyperspectral analysis framework was developed to investigate both effective wavelength selection and model-dependent spectral response mechanisms.

Development of a Compact Automatic Sorting Machine for Kazakhstani Apple Varieties Based on Computer Vision →

This article presents the development and experimental study of a compact, automated apple sorting machine based on computer vision, indirect fruit weight estimation, and color grading. It is designed for use in small and medium-sized farms, where the use of industrial lines is limited by high cost and complex maintenance. The proposed machine enables non-destructive assessing and sorting of apples in a continuous flow mode without the use of mechanical weighing, includes a fruit feeding and positioning module, a computer vision system, an image processing unit, and a sorting actuator synchronized with the conveyor movement. Fruit weight and color assessment is based on visual geometric parameters (diameter, fruit height, projected area, and the proportion of surface color), extracted from digital images, followed by classification by product category using regression models. As part of the experimental study, a correlation analysis was conducted between the actual weight of apples and their geometric parameters for five varieties typical of Kazakhstan. It was shown that the projected fruit area exhibits the most stable correlation with weight, justifying its use as the primary predictor in constructing a regression model for indirect weight estimation. An assessment of the accuracy of apple classification by product categories was carried out and the influence of conveyor speed on the stability and correctness of sorting was observed. The experimental results with a total 1250 apples of five varieties - Aport Alexander, Sinap Almaty, Kazakhski Yubileinyi, Ainur, and Nursat indicated that the optimal operating mode for the machine is an apple transport speed of 0.16 m/s. In this mode, sorting throughput is approximately 400 kg/hour, with an average accuracy of 92% for the automatic classification in accordance with GOST requirements. These results confirm that the proposed approach provides sufficient real-time sorting accuracy with a simple machine design. The machine can be used as a standalone sorting solution, as well as a base platform for further expansion of functionality by integrating surface defect assessment and grade identification modules.

Why it matches plant phenotyping methods果実の形態・色・重量を画像から推定し分類するコンピュータビジョン方式と装置の開発が中心であり、単なる品質測定ではなく、再利用可能な植物器官形質の取得・推定法を提示している。

abstractThis article presents the development and experimental study of a compact, automated apple sorting machine based on computer vision, indirect fruit weight estimation, and color grading.

Multi-Scale Feature Rectification for Crop Leaf Disease Segmentation in Complex Scenarios →

Crop leaf disease segmentation in complex natural environments remains challenging because lesion regions often exhibit substantial scale variation, blurred boundaries, and severe background interference. To address these issues, this study proposes a Multi-Scale Feature Rectification Network (MFR-Net) for crop leaf disease segmentation. The proposed network adopts an EfficientNetV2-S-based encoder to extract hierarchical features, incorporates a hybrid attention mechanism to enhance lesion-sensitive spatial and channel representations, introduces a Cross-Window Atrous Spatial Pyramid Pooling (CWASPP) module to strengthen multi-scale contextual modeling, and employs a Feature Rectification Module (FRM) in the decoder to alleviate semantic inconsistency during cross-level feature fusion. Experiments on a Kaggle-derived benchmark constructed from the unaugmented data folder of the public Leaf Disease Segmentation Dataset, containing 588 diseased-leaf images and 588 corresponding binary lesion masks, showed that MFR-Net achieved the highest mIoU of 74.27% and the highest Recall of 87.61% among the compared methods, and maintained competitive Dice performance (84.25%) with 25.10 M parameters and 37.55 G FLOPs. Ablation results further confirmed the effectiveness of the proposed design, with CWASPP providing the most notable individual contribution. Additional experiments were conducted on an independent Apple Leaf Dataset comprising 3197 image–mask pairs, collected under mixed controlled and natural field-like imaging conditions. The results showed competitive performance under a different data distribution, and robustness evaluation further verified stable performance under severe noise, blur, darkness, and contrast variation. All experiments were implemented in PyTorch 2.11.0 (CUDA 12.8) on a workstation equipped with an NVIDIA GeForce RTX 4060 Ti GPU (8 GB). These results indicate that MFR-Net provides an effective and robust solution for crop leaf disease segmentation in complex agricultural scenarios.

Why it matches plant phenotyping methods病斑領域を画像から抽出するセグメンテーション手法を開発し、複数データセット、アブレーション、ノイズ等への頑健性で検証しており、植物病害状態の表現型取得が中心である。

abstractAblation results further confirmed the effectiveness of the proposed design

Phenological growth phases of dwarf apple trees: coding and description based on the BBCH scale. →

Background The BBCH scale (Biologische Bundesanstalt, Bundessortenamt und Chemieindustrie) is a fundamental tool for standardizing plant phenological observations. As apple production shifts towards high-density dwarf orchard systems to enhance yield and facilitate mechanization, a significant gap exists in a unified phenological description framework for dwarfing rootstock apple trees, particularly under desert climate conditions. This study aims to systematically characterize the phenological growth stages of dwarf-rootstock apple using an extended BBCH scale to support precision orchard management. Result In our study, phenological monitoring was carried out systematically through the utilization of optical observation equipment and manual observations throughout the entire growth cycles. The extended BBCH scale with a three-digit coding system was applied, where the first digit indicates the principal growth stage (0-9), the second digit represents a mesostage (1 for spring, 2 for autumn growth), and the third digit describes secondary stages (0-9). We described and illustrated eight main growth stages: bud development (stage 0), leaf development (stage 1), shoot development (stage 3), inflorescence emergence (stage 5), flowering (stage 6), fruit development (stage 7), fruit ripening period (stage 8), and senescence and beginning of dormancy (stage 9). A total of 43 secondary stages were defined. Crucially, mesostages were introduced to differentiate the distinct spring and autumn vegetative growth flushes characteristic of the bimodal growth pattern observed under desert conditions. Conclusions The developed three-digit BBCH scale offers a standardized and refined framework for monitoring apple phenology in high-density dwarf orchards. It serves as a vital tool for optimizing the timing of key agronomic practices like irrigation, fertilization, pruning, and pest control, thereby supporting the intelligent implementation of precision agriculture. This framework effectively standardizes phenological observations across diverse environments, laying a foundation for improved yield, fruit quality, and sustainable orchard management.

Why it matches plant phenotyping methodsリンゴの生育段階を標準化・細分類するBBCHスケールを開発し、光学観察と手動観察によるフェノロジー取得を中心的に扱っているため、植物フェノタイピング手法として含める。

abstractThe BBCH scale (Biologische Bundesanstalt, Bundessortenamt und Chemieindustrie) is a fundamental tool for standardizing plant phenological observations.

Airborne ion charge kinetics reveal circadian-range periodicity in detached plant fruits: a continuous, noncontact, and noninvasive measurement system →

Abstract Background Understanding temporal regulation in plant systems requires measurement approaches that capture physiological kinetics with minimal tissue perturbation. Here, we present a magnetically levitated electrode ionization chamber (MALIC) system for monitoring airborne ion charge kinetics in a continuous, noncontact, and noninvasive manner, enabling the continuous physicochemical observation of plant-derived signals without genetic modification or optical readouts. Results Using detached fruits of two apple cultivars (Yoko and Akibae), Lomb-Scargle periodogram analysis identified dominant periodic components within the circadian range (~ 24–25 h) in airborne ion charge. Phase drift analysis revealed cultivar-dependent differences in temporal stability, with Yoko exhibiting relatively consistent peak timing across successive cycles, whereas Akibae showed greater variability. These differences were further supported by phase coherence analysis, which demonstrated the tighter clustering of phase values in Yoko compared with Akibae. Quantitative analysis showed that Akibae exhibited larger amplitude and a higher coefficient of variation, indicating greater relative variability. Because the MALIC system detects net ion-related signals in the surrounding air rather than intracellular processes directly, the observed oscillations should be interpreted as a proxy for integrated physiological activity. These rhythms are consistent with temporally organized physiological processes but do not establish a direct link to endogenous circadian clock mechanisms. Conclusions The MALIC system enables the continuous, noncontact, and noninvasive measurement of airborne ion charge kinetics exhibiting reproducible circadian-range periodicity in detached plant tissues. This work establishes airborne ion charge as a previously unrecognized temporal signal at the plant–environment interface. Rather than replacing established circadian assays, the MALIC system should be considered a complementary approach that captures signals distinct from transcriptional and photosynthetic readouts. Further validation across species, cultivars, and environmental conditions, together with integrated environmental, molecular, and physiological measurements, will be essential to determine whether airborne ion charge kinetics can serve as reliable indicators of endogenous biological rhythms in plants.

Why it matches plant phenotyping methodsMALICシステムによる植物由来の空中イオン荷電動態を、非接触・連続的に測定する手法の提示と、リンゴ果実での技術的適用・解析が中心である。

abstractHere, we present a magnetically levitated electrode ionization chamber (MALIC) system for monitoring airborne ion charge kinetics in a continuous, noncontact, and noninvasive manner

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.

Deep convolutional models for robust multi-crop disease recognition in real-world conditions. →

Crop diseases significantly reduce agricultural output and are a serious problem, especially in the parts of the world where diagnostic experts are not readily available. Deep learning has recently shown us that it is possible for a computer to identify plant diseases directly from images of the leaves. Nevertheless, to make such solutions available on the web or mobile devices one has to really think about how heavy the calculations will be, how easy the user interface should be, and also the limit on the data used. Here is a paper on a web-based applied deep learning system for disease detection in multiple crops. The system detects disease in eight crops Apple, Banana, Grape, Mango, Cauliflower, Tomato, Potato, and Corn with each crop having several disease classes and healthy samples. Three transfer-learning-based CNN architectures MobileNetV3, EfficientNetB4, and ResNet50 were compared for classification performance on the public datasets collected from PlantVillage, Kaggle, and Mendeley. Considering class-wise accuracy, prediction time, and deployment scenarios, MobileNetV3 was picked as the main model to be integrated into the system. To compensate for the differences in image quality often found in pictures taken by users, an optional super-resolution preprocessing step with Real-ESRGAN is added and quantitatively assessed. Disease prediction with spectral activation maps (Grad-CAM) enhances the model's interpretability by highlighting image areas where the disease is detected. The resulting model is embedded in a multilingual Progressive Web Application (PWA). The platform enables users to submit their crop images and receive predicted disease names and treatment options, which are generated by a Large Language Model (LLM) using structured disease metadata. The research acknowledges dataset bias and limitations in extrapolating from curated datasets to the general real-world setting although it reports very good performance of the method on the test sets. In summary, the system proposed here is intended as a practical digital agriculture decision-support tool that demonstrates deployment feasibility and raises a few issues for future validation at the field level and improvement.

Why it matches plant phenotyping methods葉画像から植物病害を推定するCNN手法の比較・前処理評価・実装を中心とした研究であり、植物の病害状態を直接評価するため、植物フェノタイピング手法として中心的です。

abstractDeep learning has recently shown us that it is possible for a computer to identify plant diseases directly from images of the leaves.

Research on Branch Recognition and Pruning Method for Dormant Apple Trees Based on Neural Radiance Fields and PointNeXt →

Abstract To address the problem of fine branch identification and pruning decision for dormant apple trees, this study proposes a 3D point cloud branch recognition method integrating Neural Radiance Fields (NeRF) and the PointNeXt network. This method employs the neural radiance field theory to construct a point cloud model of apple trees, achieving fine detail representation and providing a high-precision, high-standard dataset for subsequent branch pruning experiments. First, a panoramic video is captured by circling the fruit tree, and a multi-view image sequence is obtained through frame sampling. Subsequently, the Structure from Motion (SfM) algorithm is employed for sparse reconstruction to recover the pose information of the images. On this basis, a neural radiance field model is trained. Hierarchical sampling is performed using ray casting, and the sampled points, combined with positional encoding, are fed into a multi-layer perceptron (MLP). The radiance field is then generated via volume rendering, from which a high-fidelity 3D point cloud model of the fruit tree is derived. Finally, the point cloud is processed using the PointNeXt semantic segmentation network to achieve the identification and segmentation of branches to be pruned and branches to be retained. To verify the effectiveness of the method, this study reconstructed point cloud models of dormant apple trees and selected 10 of them for experimental analysis. The algorithm achieved an average overall recognition accuracy of 75.15% and an average false negative rate (FNR) of 24.85%. The experimental results demonstrate that the proposed method constructs a 3D point cloud model with multi-scale, multi-modal, and high-precision phenotypic information at a relatively low cost. It not only overcomes the limitations of traditional 3D reconstruction methods, such as insufficient point cloud accuracy and difficulty in accurately identifying thin branches, but also effectively mitigates the high misrecognition rate observed in conventional branch recognition approaches. This provides technical support for unmanned agricultural machinery pruning in orchards and holds significant implications for achieving precision agriculture and sustainable development.

Why it matches plant phenotyping methodsNeRFとPointNeXtを用いてリンゴ樹の3D点群を構築し、剪定対象枝を認識・分割する手法が研究の中心であり、植物の形態・構造状態を直接推定して性能評価している。

abstractthis study proposes a 3D point cloud branch recognition method integrating Neural Radiance Fields (NeRF) and the PointNeXt network.

Early apple moldy core classification via multi-modal sensing and SE-ResNet18. →

Apple moldy core disease is a major pathogenic disease that severely degrades the postharvest quality of apples, and its early internal lesions cannot be directly identified through visual appearance observation. To realize efficient and non-destructive early diagnosis, this study proposes a multimodal image coding method fusing Visible-Near Infrared Spectroscopy (Vis-NIR) and Electronic Nose (E-nose) data, which is combined with the SE-ResNet18 deep learning model for disease classification. By virtue of coding techniques including Gramian Angular Field (GAF), Markov Transition Field (MTF), and Recurrence Plot (RP), one-dimensional time-series and spectral data were converted into image representations, so as to visualize their spatiotemporal patterns and enhance subtle disease-related features. On this basis, a two-branch SE-ResNet18 model based on the channel attention mechanism was constructed to improve the feature representation capability and achieve effective modal fusion. Experimental results show that the multimodal fusion model achieves a classification accuracy of 95.93%, which is significantly superior to single-modal methods, thus verifying the effectiveness of multi-source information complementarity. Ablation experiments further indicate that the SE attention module plays a crucial role in feature calibration and modal balance. Information entropy analysis reveals that the proposed method effectively enhances the discriminability of information during the feature extraction process. This study provides a solution with a clear theoretical basis and reliable performance for the non-destructive detection of early diseases in agricultural products, which has favorable application prospects and popularization potential.

Why it matches plant phenotyping methodsリンゴ果実の内部病変という植物器官の病態を対象に、Vis-NIR・E-noseのマルチモーダルセンシングと深層学習による非破壊分類法を開発・評価しており、表現型取得手法が中心である。

abstractTo realize efficient and non-destructive early diagnosis, this study proposes a multimodal image coding method fusing Visible-Near Infrared Spectroscopy (Vis-NIR) and Electronic Nose (E-nose) data, which is combined with the SE-ResNet18 deep learning model for disease classification.

Enhancing precision harvesting in smart orchards: a light-weight neural network for apple maturity detection →

Introduction Deep learning-based apple maturity detection supports precise management in smart agriculture. However, deployment on resource-constrained edge devices requires minimizing network weights while ensuring accuracy, a challenge compounded by inclement weather and dense fruit clustering in orchard environments. Methods To address these challenges, we propose HRLN-YOLO, a lightweight high-efficiency apple maturity detection model that minimizes network weights while ensuring accuracy. Specifically, we design a lightweight backbone HGBackbone to enhance feature extraction and accelerate inference, construct an enhanced neck module RCF_Neck to improve multi-scale feature fusion under occlusion, develop a lightweight detection head LADH-Head to alleviate task conflicts with minimal computational cost, and introduce NWD-Loss to improve localization stability for small-scale targets. Results Experiments on the Orchard Apple Maturity Dataset demonstrate that HRLN-YOLO improves mAP@0.5 by 1.7% over the YOLO11n baseline while reducing parameters by 37.3% and computational complexity by 34.9%. Discussion The core contribution of this study lies in minimizing network weights while ensuring detection accuracy, providing a practical solution for edge deployment in smart orchard automated harvesting.

Why it matches plant phenotyping methodsリンゴ果実の成熟状態を推定する軽量画像認識モデルの開発と性能評価が中心であり、単なる収穫対象の位置検出ではなく、植物器官の状態を測定する方法である。

abstractwe propose HRLN-YOLO, a lightweight high-efficiency apple maturity detection model that minimizes network weights while ensuring accuracy.

Hybrid Optimized AgriShieldNet Framework for an Interpretable Multi‐Crop Leaf Disease Classification →

ABSTRACT Early detection of plant leaf diseases is essential for improving crop productivity, reducing financial losses, and enhancing sustainability in modern agricultural practices. However, most automated systems struggle to accurately capture diverse morphological variations of plant leaf diseases across multiple crop types, while manual inspection methods are time‐consuming, labor‐intensive, and prone to human error. To address these challenges, this paper proposes Hy‐OptiASNet, a hybrid optimized deep learning framework for interpretable multi‐crop leaf disease classification. The proposed framework utilizes EfficientNet‐B0 for robust spatial feature extraction; a Long Short‐Term Memory (LSTM) network is employed to model sequential feature dependencies and improve structured feature learning, followed by a Morphology‐Aware Vision Embedding (MAVE) module to enhance representation of fine‐grained morphological disease patterns. In addition, a novel Sandpiper Optimization Algorithm (SPOA) is incorporated for feature refinement and hyperparameter optimization, thereby improving generalization capability and classification stability. To enhance interpretability, visualization techniques are used to highlight biologically relevant disease regions, enabling improved understanding of model decisions. Extensive experiments were conducted on multi‐crop datasets consisting of apple, tomato, and grape leaf images. The experimental results demonstrate that the proposed Hy‐OptiASNet framework achieves classification accuracies of 98.98%, 98.21%, and 98.35% for apple, tomato, and grape datasets, respectively, outperforming several existing state‐of‐the‐art methods. These findings indicate that the proposed framework provides an effective and reliable solution for precision agriculture and real‐world plant disease monitoring applications.

Why it matches plant phenotyping methods植物葉の病徴を画像から分類する新規深層学習フレームワークの開発・評価が中心であり、植物の疾病状態を直接推定するため、植物フェノタイピング手法に該当します。

abstractthis paper proposes Hy‐OptiASNet, a hybrid optimized deep learning framework for interpretable multi‐crop leaf disease classification

Inversion of nitrogen content in apple leaves using an explainable PSO-CNN model. →

Introduction The integration of hyperspectral technology with machine learning and deep learning algorithms offers an effective method for accurately and non-destructively estimating the percentage of nitrogen in apple tree leaves, as well as for rapid nutrient diagnosis. Methods This study was conducted in apple orchards in Qixia City, Shandong Province, where hyperspectral data were collected from Red Fuji apple trees during the new-shoot-stop-growing stage (NSS) and the autumn-shoot-stop-growing stage (ASS). Following hyperspectral preprocessing and characteristic wavelength selection, regression models-including random forest (RF), support vector machine (SVM), convolutional neural network (CNN), and particle swarm optimization convolutional neural network (PSO-CNN)-were developed and compared. Results The results showed that CNN significantly outperformed RF and SVM, and that PSO-CNN further improved prediction performance. The PSO-CNN model achieved an R² of 0.886 on the training set and an R² of 0.774, an RMSE (%) of 0.095, and an RPD of 2.086 on the test set. Validation using samples from different phenological stages demonstrated acceptable prediction accuracy ( R 2 = 0.625, RMSE = 0.173, RPD = 1.529), with uniformly distributed errors and no systematic bias, indicating good generalization capability and stability. SHAP analysis revealed that the PSO-CNN model primarily relied on the near-infrared and short-wave infrared bands, where reflectance was negatively correlated with nitrogen content. Discussion These spectral regions are closely associated with leaf biochemical structure, suggesting that the model predicts nitrogen content by capturing spectral information related to leaf biochemical characteristics. Overall, the PSO-CNN model improves nitrogen prediction performance by expanding the hyperparameter search space while preserving the CNN architecture, enabling rapid and accurate nutrient diagnosis in apple leaves.

Why it matches plant phenotyping methodsリンゴ葉の窒素含量という植物形質を、ハイパースペクトル計測とPSO-CNNで推定する手法を開発し、複数モデル比較と異なる生育段階での検証を行っており、表現型取得・推定が研究の中心である。

abstracthyperspectral technology with machine learning and deep learning algorithms offers an effective method for accurately and non-destructively estimating the percentage of nitrogen in apple tree leaves

Apple Plant Disease Detection System using Leaf Images →

The cultivation of apples is affected by various apple plant diseases. These diseases, if not identified and treated on time, may lead to considerable losses in yield. Early detection is highly essential in order to provide early warnings to farmers and to help in identifying diseases at an early stage so that further action can be done to prevent the spread of disease as these diseases cannot be identified through naked eyes in their early stages. This leads to less wastage of yield. This paper proposes a comparison among the deep learning models such as LetNet, AlexNet, VGG, Resnet, Inception Net, and DensNet, for the efficient classification of leaf diseases of the apple plant. The model is trained on the Plant Village Dataset (Updated) taken from kaggle, which contains both healthy and diseased leaf images of apple plants. In this, images go through various preprocessing techniques like resizing, normalizing, and augmenting images in order to increase the robustness of the model. AlexNet attained the maximum classification accuracy in initial trials but had the largest number of parameters, didn't use modern regularization, and hence got at risk of overfitting at some early stage. The paper improved their performance by using a hybrid architecture which consisted of MobileNetV3 and ResNet50 because MobileNetV3 offered efficient extraction of features with little computational expense. It was further complemented by the depth features offered by ResNet50. The main aim for proceeding with the idea of hybrid architecture was not only to improve generalization but also to prevent overfitting. The hybrid model is implemented using streamlit. This web interface allows the users to upload images of leaves and get real-time results predicting whether the leaves are affected by a disease or not. The system demonstrates high classification accuracy and effective differentiation among visually similar diseases. However, the model's performance in terms of empirical data analysis is influenced by dataset quality, computational resource demands, and its limited ability to generalize in the presence of sparse data. Despite these challenges, the proposed solution provides a scalable and accessible tool to assist farmers and agricultural experts in early disease detection and management.

Why it matches plant phenotyping methodsリンゴ葉の画像から病徴・病害状態を推定する画像ベースの植物フェノタイピング手法を開発・比較しており、分類モデルと実装が研究の中心である。

abstractThis paper proposes a comparison among the deep learning models such as LetNet, AlexNet, VGG, Resnet, Inception Net, and DensNet, for the efficient classification of leaf diseases of the apple plant.

Simulation-Driven Spatial Frequency Domain Imaging and Deep Learning for Subsurface Fruit Bruise Discrimination. →

Conventional spatial frequency domain imaging (SFDI) based optical property inversion is inefficient, while deep learning methods suffer from heavy reliance on large-scale real datasets. To address this contradiction, a simulation-driven approach for subsurface fruit bruise discrimination was proposed. An SFDI simulation environment was built with Blender to generate 800 paired datasets of diffuse reflectance images and optical transport coefficients, overcoming the high cost and long cycle of real dataset acquisition. We designed the CBAM-GAN-U-Net model and adopted surface profile correction in the prediction method to eliminate curved surface-induced non-planar distortion, with the whole method validated on liquid phantoms, green apples and crown pears. This prediction method achieved high accuracy in predicting the reduced scattering coefficient μ s ', with NMAE of 0.021 ± 0.007 (phantoms), 0.039 ± 0.012 (severely bruised green apples) and 0.044 ± 0.015 (severely bruised crown pears), outperforming U-Net and GANPOP. Based on the predicted μ s ', a discrimination strategy combining coefficient of variation, mean ratio and receiver operating characteristic (ROC) curve analysis was adopted, attaining 100% accuracy for non-bruised/bruised fruit discrimination, with misclassification rates of 6% (green apples) and 8% (crown pears) for mild/severe bruise differentiation. This method enables accurate subsurface fruit bruise detection, providing a reliable technical solution for the fruit and vegetable industry and helping reduce postharvest supply chain losses.

Why it matches plant phenotyping methodsSFDIと深層学習による果実内部の打撲状態・重症度の画像推定手法を開発し、ファントムと果実で検証しており、植物（果実）の状態取得が中心である。

abstracta simulation-driven approach for subsurface fruit bruise discrimination was proposed.

EffcientNetB0-based Deep Learning Approach for Early Plant Disease Detection using Leaf Image Classification →

Plant diseases can damage crops and reduce their growth, which often results in economic problems in agriculture. Detecting these diseases at an early stage is essential to protect crop health and improve productivity. In this work, we developed an automated plant disease identification system using deep learning and leaf images from crops such as corn, grapes, and apples. Our introduced model is based on EfficientNetB0, which showed outstanding performance in classifying multiple disease categories. To make the system more reliable, data augmentation techniques were used to manage different lighting condition in lighting, angles, and backgrounds. The model obtained an accuracy of 97.92%, outperforming other architectures like InceptionV3, MobileNetV2, DenseNet121, Xception, and VGG16 with a lesser accuracy of 83.33%, 78.33%, 77.08%, 75%, 73.33% respectively. Performance measures like precision, recall, and F1-score confirmed its strong performance. The confusion matrix further showed that the model effectively distinguishes between healthy and diseased leaves. This approach provides a fast and accurate solution for real-time disease detection. Overall, the proposed system can support farmers in taking timely action and promoting sustainable agricultural practices.

Why it matches plant phenotyping methods葉画像から植物の病害状態を推定する深層学習手法を開発・比較評価しており、植物表現型の取得・分類が研究の中心であるため。

abstractwe developed an automated plant disease identification system using deep learning and leaf images

Vision transformer-based approach for plant leaf disease classification with multi-patch selection →

Plant diseases pose significant challenges to global agricultural productivity, requiring early and accurate detection to mitigate their impact. This study introduces a Vision Transformer (ViT)-based framework for classifying apple leaf diseases, utilizing a novel multi-patch selection approach to strike a balance between feature extraction and computational efficiency. Leaf images were preprocessed to ensure compatibility with the ViT framework and divided into patches of varying sizes (32 × 32, 16 × 16, and 8 × 8), enabling the model to capture both local and global features for classifying four categories: Apple Scab, Black Rot, Cedar Rust, and Healthy Leaves. The proposed Plant Leaf Disease Vision Transformer (PLD-ViT) framework achieves superior classification performance with 97.76% validation accuracy, 95.34% precision, 95.11% recall, and 95.06% F1-score using 16 × 16 patch configuration, significantly outperforming ResNet-50 (96.75% validation accuracy, 95.11% precision, 93.42% recall, 95.31% F1-score) and Swin Transformer (96.79% validation accuracy, 95.76% precision, 95.94% recall, 95.89% F1-score) while maintaining computational efficiency (1.0× GFLOPs vs 3.8× and 4.5× respectively). The model robustly classifies distinct categories but faces challenges distinguishing visually similar groups, such as Cedar Rust and Healthy Leaves. Despite its strengths, the model has limitations, including its reliance on controlled datasets and the computational demands associated with smaller patch sizes. This ViT-based framework offers a practical and scalable solution for precision agriculture, laying the groundwork for accessible tools that support sustainable farming practices and promote global food security.

Why it matches plant phenotyping methods植物葉の病害状態を画像から分類するViTベース手法を開発し、複数モデルとの性能比較・検証を行っており、病害表現型の取得・推定が研究の中心である。

abstractThis study introduces a Vision Transformer (ViT)-based framework for classifying apple leaf diseases, utilizing a novel multi-patch selection approach

Measurement of Fruit Diameter Using RGB‐D Cameras for the Purpose of Fruit Growth Assessment →

ABSTRACT This article describes a method for measuring the diameter of fruits with a near spherical shape using an RGB‐D camera in order to investigate fruit enlargement at different times of the year. In general, depth‐based measurement methods, when considering the diameter of an object with a near‐spherical shape, the diameter can be calculated by obtaining the Euclidean distance from the coordinates of the object's sides or by using the depth at the center of the object and the size of the object on the RGB image. However, it is difficult to accurately determine the diameter of an object because of errors in the calculated results due to the perspective projection of a general camera. Therefore, this study proposes a method to measure the diameter of fruits that have a shape similar to a sphere. Although this study focuses on young apple fruits, the proposed method can be applied to other agricultural crops, as well as to objects that are similar to spheres. In addition, we have also studied a correction that takes into account the rotation of the object so that the method can be applied to objects with circular cross‐sections.

Why it matches plant phenotyping methodsRGB-Dカメラを用いて果実径を測定する手法を提案・補正しており、植物形質の取得方法が研究の中心である。

abstractThis article describes a method for measuring the diameter of fruits with a near spherical shape using an RGB‐D camera

Robust apple leaf disease diagnosis for sustainable horticulture: overcoming background noise with a multilayer transformer-based approach. →

Apples are one of the most economically significant and widely cultivated fruit crops worldwide, contributing substantially to food security and the horticultural economy. However, their production is frequently compromised by diseases such as Alternaria leaf spot, Apple Mosaic, Powdery Mildew, and Apple Scab, leading to significant yield losses and increased dependence on chemical control. Timely and accurate disease diagnosis is critical to minimize crop damage, reduce pesticide usage, and promote sustainable horticultural practices. This study proposes the Multilayer Transformer-based Apple Disease Classification (MTADC) model, an advanced deep learning framework designed for early and robust identification of apple leaf diseases. MTADC employs a two-stage learning approach: global feature extraction using a transformer encoder, followed by class-specific mapping through a refined classification head. By incorporating a self-attention mechanism, the model effectively suppresses background noise and enhances feature discrimination, even under variable field conditions. Unlike existing models that require well-constrained, high-quality images captured under ideal lighting and angles, MTADC is designed for deployment in real-world field conditions, allowing disease detection from diverse and unconstrained field images commonly captured by farmers. Experiments on a curated dataset comprising publicly available and field-acquired images demonstrate that MTADC achieves a classification accuracy of 96.3%, outperforming conventional convolutional models. These results highlight the model’s robustness, scalability, and potential to be an accessible tool for digital plant health monitoring and precision horticulture.

Why it matches plant phenotyping methodsリンゴ葉の病害状態を画像から推定する深層学習モデルを開発し、背景ノイズや圃場条件への頑健性を評価しており、植物フェノタイピング手法が研究の中心である。

abstractThis study proposes the Multilayer Transformer-based Apple Disease Classification (MTADC) model, an advanced deep learning framework designed for early and robust identification of apple leaf diseases.

Multi-Class Classification of Plant Leaf Diseases Using Feature Fusion of Deep Convolutional Neural Network and Segmentation Techniques →

ABSTRACT A multi-stage approach for plant leaf disease classification using Convolutional Neural Networks (CNN) and image segmentation is presented. The system identifies plant types and classifies diseases from leaf images. Preprocessing includes noise removal and image enhancement, followed by K-means clustering for segmentation. A hybrid CNN model performs accurate classification of plant species and disease type, and suitable fertilizer recommendations are provided. The method achieves accuracies of 99.8%, 96.5%, and 98.3% on apple, tomato, and grape datasets. The results show improved accuracy and robustness, making the system effective for practical agricultural applications. INDEX TERMS Convolutional Neural Network (CNN), Image Segmentation, K-means clustering, Plant Leaf Diseases.

Why it matches plant phenotyping methods葉画像から病害状態を直接推定するCNN・画像セグメンテーシ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手法として該当する。

abstractA multi-stage approach for plant leaf disease classification using Convolutional Neural Networks (CNN) and image segmentation is presented.

Photosynthetic traits and canopy-level thermal imaging to assess plant-water relations in two apple cultivars under waterlogging and recovery conditions →

Waterlogging is an increasingly important constraint in orchard systems under climate extremes. Understanding cultivar-specific physiological responses and identifying reliable, non-invasive indicators of plant water status are essential for improving orchard management under excess soil moisture. In this study, we evaluated the physiological, hydraulic, and canopy thermal responses of two commercially important apple cultivars, ‘Fuji’ and ‘Gamhong' grafted onto M.9, under controlled waterlogging and recovery conditions. We quantified photosynthetic traits and plant hydraulic parameters—including sap flow (SF), leaf water potential ( Ψ Leaf ), and whole-plant hydraulic conductivity ( K s )—together with canopy thermal indicators, canopy temperature ( T c ), and a modified crop water stress index ( mCWSI ) calculated using empirically derived, day-specific canopy temperature references. Waterlogging significantly reduced photosynthetic performance and hydraulic function in both cultivars, but responses differed in magnitude and recovery dynamics. ‘Fuji’ exhibited greater resilience, with smaller declines and faster recovery of gas exchange and water-relation traits, whereas ‘Gamhong’ showed earlier photosynthetic limitation and delayed recovery, indicating lower tolerance to saturated soil conditions. Leaf mass per area (LMA) increased under waterlogging, reflecting constraints on leaf expansion rather than enhanced photosynthetic activity. Among the thermal indicators, mCWSI showed the strongest correlations with Ψ Leaf , stomatal conductance ( g s ), and net photosynthetic rate ( P n ), outperforming T c as an indicator of plant water status. These findings demonstrate that canopy-based thermal metrics, particularly mCWSI when interpreted alongside physiological traits, provide a robust tool for detecting cultivar-specific responses to waterlogging stress. This multi-trait framework supports cultivar selection and precision water management in orchard systems exposed to episodic flooding.

Why it matches plant phenotyping methodsキャノピー熱画像から算出したmCWSIを生理・水分状態の指標として検証し、従来のキャノピー温度と比較しているため、表現型取得・評価法が研究の中心的要素である。

abstracta modified crop water stress index ( mCWSI ) calculated using empirically derived, day-specific canopy temperature references

Geo-Referenced Factor-Graph SLAM for Orchard-Scale 3D Apple Reconstruction and Yield Estimation →

Accurate and spatially resolved yield estimation is a critical requirement for precision agriculture and orchard management. This paper presents a geometrically consistent, orchard-scale apple yield estimation framework that integrates GNSS–visual-inertial odometry (VIO) fusion, deep learning-based object detection, multi-frame tracking, three-dimensional triangulation, and incremental factor-graph optimization. Camera poses are obtained using ZED GNSS–VIO fusion and subsequently refined using an iSAM2-based nonlinear smoothing approach that incorporates strong relative-motion constraints and soft global ENU (East-North-Up) translation priors. Apples are detected using a YOLO-based model and associated across frames via CoTracker3, enabling robust multi-view landmark reconstruction. Reprojection factors and landmark priors are incorporated into a unified nonlinear factor graph to jointly optimize camera trajectories and 3D apple positions. The reconstructed apples are spatially aggregated into a grid-based mass map, where individual fruit volumes are estimated assuming spherical geometry and converted to mass using density models. The resulting ENU-referenced yield plot provides a structured representation of orchard production variability. Experimental results demonstrate significant reductions in reprojection error after optimization and improved global consistency of the trajectory, leading to stable and spatially coherent 3D reconstructions. The proposed pipeline bridges perception, geometry, and optimization, providing a scalable solution for orchard-scale yield mapping and decision support in precision agriculture.

Why it matches plant phenotyping methods果実の三次元再構成から体積・質量・収量を推定する画像・計算パイプラインが研究の中心であり、植物器官の形態および収量形質を技術的に抽出している。

abstractThis paper presents a geometrically consistent, orchard-scale apple yield estimation framework that integrates GNSS–visual-inertial odometry (VIO) fusion, deep learning-based object detection, multi-frame tracking, three-dimensional triangulation, and incremental factor-graph optimization.

Early Apple Yield Prediction Based on Flowering Stage Image Thinning Simulation Characteristics. →

The existing fruit tree yield prediction methods mainly rely on fruit period images or long-term meteorological and soil data, which make it difficult to meet the needs of early yield prediction. In addition, the flowering period images contain complex spatial distribution and severe overlap between flowers, which makes it challenging to directly extract stable structural indicators related to yield. Most existing research has focused on simple statistical indicators such as the number of flowers, while the spatial clustering structure of flowers and their relationship with yield have not been fully explored. Therefore, this article proposes an early apple yield prediction based on flowering stage image thinning simulation characteristics. In this study, blossom images and fruit maturity yield data from 100 apple trees were collected, with flower mask images extracted through standardized image processing. First, the traditional DBSCAN clustering algorithm was enhanced by integrating a KDTree acceleration structure and an adaptive multi-scale mechanism, forming the adaptive multi-scale clustering algorithm (AMS-DBSCAN) to achieve efficient identification of flower clusters and individual flowers. Based on this, two flower thinning simulation strategies based on density and spatial uniformity were designed to model artificial thinning rules and construct multi-dimensional, interpretable phenotypic features. Then, the original statistical features were fused with strategy-generated features and optimized using Lasso. We compared multiple models including XGBoost, BPNN, and SVR for yield prediction. The experimental results showed that XGBoost achieved good predictive performance under the hybrid feature set (R 2 = 0.856, RMSE = 3.098), which was further improved to R 2 = 0.900 after feature optimization with Lasso. The results demonstrate that the proposed method enables reliable early yield estimation, providing a new reference for precision management and early decision-making in fruit tree cultivation.

Why it matches plant phenotyping methods花画像から花群・個体を抽出し、間引きシミュレーション由来の解釈可能な表現型特徴を構築する画像解析手法が研究の中心であり、収量推定に技術的に応用・評価されている。

abstractflower mask images extracted through standardized image processing

A Reproducible Methodology for 3D Tree-Structure Mensuration and Risk-Oriented Decision Support: Integrating SfM–MVS, Field Referencing, and Rule-Based TRAQ/ALARP Logic →

This manuscript presents a transferable and reproducible methodology for quantitative 3D tree-structure mensuration and transparent, rule-based decision support for tree risk management. The workflow integrates (i) Structure-from-Motion/Multi-View Stereo (SfM–MVS) reconstruction from multi-view imagery, (ii) independent referencing to ensure metric scaling and a consistent local frame, and (iii) point cloud analytics to derive branch-level geometric descriptors (e.g., base diameter, length, inclination, slenderness, and projected reach). A clear rule-based layer operationalizes Tree Risk Assessment Qualification (TRAQ)-style risk components and As Low As Reasonably Practicable (ALARP) principles to map geometry and exposure into auditable management recommendations (e.g., monitoring intervals, pruning/weight reduction, supplemental support, and exclusion-zone planning). To provide a real-data example, the demonstration uses the public Fuji-SfM apple orchard dataset, including three neighboring trees with partially overlapping crowns for tree instance extraction and subsequent TRAQ/ALARP scenarios on an outer tree. The proposed decision layer is intentionally based on external geometry and exposure; internal decay indicators and species-specific mechanical properties (e.g., Modulus of Elasticity (MOE), Modulus of Rupture (MOR)) are outside this demonstration and should be incorporated via complementary diagnostics in operational deployments.

Why it matches plant phenotyping methodsSfM–MVSと点群解析による樹木・枝の3D形状形質抽出が中心的な方法論であり、実データでの適用も含むため。

abstractThis manuscript presents a transferable and reproducible methodology for quantitative 3D tree-structure mensuration

Research on Lightweight Apple Detection and 3D Accurate Yield Estimation for Complex Orchard Environments →

Severe foliage occlusion and dynamically changing lighting conditions in complex orchard environments pose significant challenges for visual perception systems in automated apple harvesting, including low detection accuracy, poor robustness, and insufficient real-time performance. To address these issues, this study proposes an improved lightweight detection network based on YOLOv11, named YOLO-WBL, along with a precise yield estimation algorithm based on 3D point clouds, termed CLV. The YOLO-WBL network is optimized in three aspects: (1) A C3K2_WT module integrating wavelet transform is introduced into the backbone network to enhance multi-scale feature extraction capability; (2) A weighted bidirectional feature pyramid network (BiFPN) is adopted in the neck network to improve the efficiency of multi-scale feature fusion; (3) A lightweight shared convolution separated batch normalization detection head (Detect-SCGN) is designed to significantly reduce the parameter count while maintaining accuracy. Based on this detection model, the CLV algorithm deeply integrates depth camera point cloud information through 3D coordinate mapping, irregular point cloud reconstruction, and convex hull volume calculation to achieve accurate estimation of individual fruit volume and total yield. Experimental results demonstrate that: (1) The YOLO-WBL model achieves a precision of 93.8%, recall of 79.3%, and mean average precision (mAP@0.5) of 87.2% on the apple test set; (2) The model size is only 3.72 MB, a reduction of 28.87% compared to the baseline model; (3) When deployed on an NVIDIA Jetson Xavier NX edge device, its inference speed reaches 8.7 FPS, meeting real-time requirements; (4) In scenarios with an occlusion rate below 40%, the mean absolute percentage error (MAPE) of yield estimation can be controlled within 8%. Experimental validation was conducted using apple images selected from the dataset under varying lighting intensities and fruit occlusion conditions. The results demonstrate that the CLV algorithm significantly outperforms traditional average-weight-based estimation methods. This study provides an efficient, accurate, and deployable visual solution for intelligent apple harvesting and yield estimation in complex orchard environments, offering practical reference value for advancing smart orchard production.

Why it matches plant phenotyping methodsリンゴ果実の検出と3D点群による個別果実体積・総収量推定を開発し、精度・速度・遮蔽条件下で検証しており、植物形質取得が中心的な方法論的貢献である。

abstractthis study proposes an improved lightweight detection network based on YOLOv11, named YOLO-WBL, along with a precise yield estimation algorithm based on 3D point clouds, termed CLV.

YOLOv9s-multi: orientation-based fruit selection for robotic apple thinning →

Accurate detection and orientation estimation of immature apples are crucial for effective thinning decisions in robotic apple thinning. Existing research either relies on computationally expensive RGB-D approaches with 3D geometric fitting to estimate orientation and size or requires multiple separate models for thinning decision, limiting their real-time performance on robotic platforms. To address these issues, a multi-task model, YOLOv9s-Multi, is proposed. First, this model integrates segmentation and keypoint heads to perform instance segmentation of immature apples and detect their calyx keypoint positions. Second, the detection head employs an efficient and lightweight Depthwise Convolution module (DWConv) to reduce model parameters while accurately capturing spatial features across channels. Finally, the orientation is derived from the segmentation centroid and calyx keypoint, enabling pixel-based fruit selection for thinning decision-making. This model is evaluated on a self-developed dataset that divides immature apples based on developmental stage: Flower-Retained Stage (FR-Stage) and Fruit-Visible Stage (FV-Stage). Results show that instance segmentation and keypoint AP@0.5 for FV-Stage are 89.3% and 86.4%, respectively, while for FR-Stage they are 71.6% and 79.8%. The model further achieves prediction accuracies of 92.80% (FV-Stage) and 72.59% (FR-Stage) within an acceptable error of 30 °. The pixel-based fruit selection method achieves 74.00% and 70.31% selection accuracy on the test and an additional measurement dataset, respectively. Compared with the baseline YOLOv9s-seg, the number of parameters is reduced by 11.4%. In contrast to 3D fitting methods, our approach provides lower computational complexity, faster inference speed, and higher accuracy. These results demonstrate that the proposed model can efficiently estimate the orientation of immature apples and perform fruit selection in close-range scenes and complex lighting environments, which are challenging for depth cameras to handle. The code and datasets are publicly available on GitHub: https://github.com/DIANSLEE/YOLOv9s-Multi.

Why it matches plant phenotyping methods未熟リンゴのセグメンテーション、萼点検出、重心との関係から果実の向きという器官形質を推定する画像解析手法が研究の中心であり、精度評価とデータセット検証も行っているため。

abstracta multi-task model, YOLOv9s-Multi, is proposed.

Interpretable crop pest and disease identification based on comparative concept tree. →

Deep learning provides new methods for crop pest and disease identification and control, offering unique advantages in terms of recognition accuracy and efficiency. However, deep learning models generally lack interpretability, and their internal decision-making processes are difficult to understand. This, to some extent, undermines users' trust in the model's predictions and hinders its large-scale application in agricultural production. Therefore, improving model transparency and interpretability has become an important research direction. To address this issue, this study proposes a novel interpretable crop pest and disease identification model, the Contrastive Prototype Tree (CPTR). The model is designed around the core structure of "concept prototypes and decision tree," which builds clear prototype matching paths for each recognition result. This enables the model to not only have strong classification capability but also provide intuitive explanations. Additionally, the study introduces the SimCLR contrastive learning framework to enhance the model's ability to express deep image features. SimCLR guides the model to learn more discriminative visual features by maximizing the similarity between positive sample pairs and minimizing the similarity between negative sample pairs, thereby improving overall recognition performance. This study evaluated the model on three datasets: AppleLeaf9, Cassava, and Cashew. The experimental results show that CPTR achieves accuracies of 83.74%, 94.80%, and 96.01% on the three datasets, representing improvements of 4.12%, 0.34%, and 0.51% compared to Prototype Tree, respectively. These results indicate that the proposed model achieves the highest accuracy across different datasets, demonstrating its effectiveness.

Why it matches plant phenotyping methods植物葉画像から病害状態を推定する解釈可能な深層学習モデルを開発し、複数データセットで性能評価しているため、植物フェノタイピング手法が中心である。

abstractthis study proposes a novel interpretable crop pest and disease identification model, the Contrastive Prototype Tree (CPTR).

MSA-MVSNet: A Cross-Scale Collaborative Attention-Based Multi-View Reconstruction Network for Orchard Tree 3D Reconstruction with Instance Segmentation for Fruit Counting →

Abstract To address the issues of detail loss and matching difficulties in fruit tree 3D reconstruction caused by complex branch–leaf morphology, fruit occlusion, and illumination variations, this paper proposes an end-to-end cross-scale collaborative attention multi-view stereo network, termed MSA-MVSNet, for high-quality 3D reconstruction of orchard trees, while integrating semantic segmentation for fruit counting. A multi-scale feature enhancement module is designed to adaptively fuse deep semantic features and shallow fine-grained details through a spatial–channel collaborative attention mechanism, thereby enhancing the network’s capability to represent multi-scale structures such as trunks, branches, and leaves. Multi-branch dilated convolutions are introduced to enlarge the receptive field, and deformable convolutions are incorporated to adaptively capture the irregular geometric shapes of fruits, improving modeling robustness. In addition, a feature matching transformer is introduced to strengthen long-range global contextual correlations within and across images via intra-attention and inter-attention mechanisms, thereby improving matching stability in low-texture and repetitive-texture regions.To validate the effectiveness of the proposed method, experiments are conducted on self-collected real orchard dataset and public benchmark datasets. The results demonstrate that MSA-MVSNet outperforms baseline models by 8.2% in terms of 3D reconstruction quality. Finally, by combining depth filtering with the semantic segmentation results of YOLOv11-Seg, a semantic-guided fruit reconstruction and counting framework is constructed. This framework achieves an overall counting F1-score of 92.8% on the self-collected dataset with varying scene sparsity and 93.5% on the public Fuji-sfm dataset, demonstrating its effectiveness and generalization capability.

Why it matches plant phenotyping methods果樹の3D再構成と果実カウントという植物形質取得を目的に、マルチビュー再構成ネットワークとセグメンテーション統合手法を開発・検証しており、フェノタイピング手法が中心である。

abstractthis paper proposes an end-to-end cross-scale collaborative attention multi-view stereo network, termed MSA-MVSNet, for high-quality 3D reconstruction of orchard trees, while integrating semantic segmentation for fruit counting.

Online detection of apple moldy core using near-infrared spectroscopy with flexible transmission tray and deep learning. →

Apple moldy core (AMC) causes substantial postharvest losses, yet early-stage infections remain difficult to detect due to the absence of visible symptoms. This study proposed an integrated, industry-ready approach that combines transmission near-infrared (NIR) spectroscopy with a custom flexible transmission tray and deep-learning classification to enable accurate, high-throughput detection of early AMC. The tray was engineered to stabilize fruit positioning, reduce ambient-light interference, and guide NIR illumination through the fruit core, yielding reproducible transmission spectra. Spectral data were preprocessed with Savitzky-Golay smoothing, standard normal variate, multiplicative scatter correction, and mean centering. The study systematically evaluated wavelength selection strategies (CARS, SCARS and SCARS combined with SPA) and developed two-class (healthy/diseased) and three-class (healthy/mild/severe) classifiers using BP, CNN, LSTM and a hybrid CNN-LSTM architecture. The CNN-LSTM model trained on SCARS-SPA-selected wavelengths achieved the best performance, with classification accuracies of 98.82% (two-class) and 97.65% (three-class). These results demonstrate that the SCARS-SPA + CNN-LSTM pipeline, together with the flexible transmission tray, provides a robust and reproducible framework for early, precise AMC detection. The proposed system is compatible with conveyor-based integration and real-time sorting, offering a practical solution to reduce economic losses and improve quality control in commercial apple supply chains.

Why it matches plant phenotyping methodsリンゴ果実の病害状態をNIR分光と深層学習で直接推定する取得・解析システムを開発し、分類性能を評価しており、病害フェノタイピング手法が中心である。

abstractThis study proposed an integrated, industry-ready approach that combines transmission near-infrared (NIR) spectroscopy with a custom flexible transmission tray and deep-learning classification to enable accurate, high-throughput detection of early AMC.

An Innovative Design for Early Disease Detection in Apple Leaf and Fruit Using a Multi-Label Classification Model and Gaussian Bounding →

Early Disease Detection (EDD) in plants is crucial for identifying infections before serious symptoms manifest, ensuring agricultural product safety, reducing chemical use, improving orchard hygiene, and providing a sustainable disease control method. The research developed a EDALMG model to improve apple disease recognition and orchard productivity by combining RGB and multispectral imaging. It segments data from Kaggle datasets for training and creates vegetation indices to highlight plant physiology. The model uses a data-level fusion approach, channel attention, and hierarchical feature extraction through ResNet for comprehensive disease features. It employs multi-label classification to identify multiple diseases simultaneously, enhancing system robustness. The Gaussian bounding method is preferred for tracking infection areas due to its precise 2D positional data usage. The performance analysis includes Loss Calculation, Accuracy Calculation, Confusion Matrix Calculation, and F1-Score Calculation.

Why it matches plant phenotyping methodsリンゴ葉・果実の感染領域と病害状態をRGB・マルチスペクトル画像から推定する分類・位置推定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 research developed a EDALMG model to improve apple disease recognition

ADA-Net: A Lightweight Model for Apple Flower Maturity Detection in Horticultural Plant Monitoring →

In modern orchards, the pollination process of apple blossoms plays a crucial role in determining both the quality and yield of the fruit. While most current studies concentrate on identifying individual apple flowers, there is limited research on assessing the developmental stages of apple flowers in dynamic and complex orchard settings. Challenges arise due to the intricate environmental factors and subtle color changes in the anthers following the maturation of the apple flowers, which complicate accurate detection. To overcome these challenges, ADA-Net, an efficient YOLOv8n-based detection model, is proposed to evaluate the maturity stages of apple flowers. First, the adaptive downsampling network module replaces the conventional downsampling convolution, which reduces the size of the convolutional kernels and groups input feature mean average precision (maps). This modification helps reduce the model’s parameter count and computational complexity, while simultaneously improving detection of small targets. In addition, inspired by the task alignment technique of the task-aligned one-stage object detection (TOOD) model, a DAD Head is employed to separate the classification from localization tasks, thus minimizing task interference and improving overall accuracy. A custom apple flower dataset is used to test the model, and the results show detection accuracies of 80.7% for mature flowers and 82.4% for immature flowers, with a total model parameter count of just 1.8 million. These results offer important insights for advancing the development of automated pollination systems in orchards.

Why it matches plant phenotyping methodsリンゴ花の成熟段階という植物状態を画像から推定する軽量検出モデルを開発し、専用データセットで精度検証しているため、フェノタイピング手法が中心である。

abstractADA-Net, an efficient YOLOv8n-based detection model, is proposed to evaluate the maturity stages of apple flowers.

Improved YOLOv8 for multi-colored apple fruit instance segmentation and 3D localization →

Robotic apple harvesting requires precise instance segmentation and 3D localization, especially for multi-colored apples under complex orchard conditions with occlusions and variable lighting. Current deep learning methods lack robustness and accuracy for such scenarios, limiting automation. This study proposes improved YOLOv8-based models to enhance segmentation of multi-colored apples, combined with a high-precision 3D localization pipeline to advance practical robotic harvesting. To address these issues, this study collected apple images in three colors from two locations, creating a dataset of 5171 images. Four enhanced YOLOv8-based models—RA-YOLO, GA-YOLO, YA-YOLO, and MCA-YOLO—were proposed for segmenting red, green, yellow, and mixed multi-colored apples. RA-YOLO integrates the GD mechanism and EMBConv structure based on EfficientNet's MBConv. GA-YOLO replaces standard convolutions with dynamic serpentine convolution and adds the P6 layer for large object detection. YA-YOLO utilizes deformable convolution (DCNv2) and introduces the new attention mechanism MPCA. MCA-YOLO combines the P6 layer, DCNv2, and EMBConv structure, merging the strengths of other models. RA-YOLO, GA-YOLO, and YA-YOLO achieved mAP values of 95.2 %, 96.4 %, and 95.4 %, respectively, for single-colored apple instance segmentation, surpassing baseline models and those in existing literature. MCA-YOLO achieved mAP values of 95.6 %, 96.6 %, and 94.6 % for single-colored apples and 95.6 % for mixed multi-colored apples. Ablation experiments validated the necessity of each module. Finally, a high-precision 3D localization and shaping pipeline was developed, achieving an average localization error of 2.636 mm and a shaping error of 0.768 mm, enabling millimeter-level localization and sub-millimeter-level shaping for apple harvesting optimization.

Why it matches plant phenotyping methodsリンゴ果実のインスタンス分割、3D位置推定、形状推定を開発・検証しており、収穫対象の単なる検出を超えて果実形状という植物器官形質を定量化する手法が中心である。

abstractThis study proposes improved YOLOv8-based models to enhance segmentation of multi-colored apples, combined with a high-precision 3D localization pipeline

Explainable AI-Guided Hyperspectral Feature Selection in Fruit Quality Assessment and Spatial Visualization. →

The integration of hyperspectral imaging (HSI) with machine learning enables non-destructive prediction and visualization of food quality. However, multicollinearity and redundant features in spectral data can reduce model accuracy and increase computational time, emphasizing the need for key wavelength selection. In response, this study presents an inventive method combining a genetic algorithm (GA) with explainable artificial intelligence (XAI) to select key wavelengths for predicting apple dry matter content (DMC). A partial least squares regression (PLSR) model using the selected features outperformed recursive feature elimination (RFE) and competitive adaptive reweighted sampling (CARS), achieving a coefficient of determination (R 2 ) of 0.46 and a root mean squared error (RMSE) of 0.70%. The approach was further applied to hyperspectral images to visualize pixelwise DMC distribution, providing spatial insights into fruit composition. Results demonstrate that integrating XAI with evolutionary feature selection offers a noninvasive, transparent, and efficient strategy for assessing and visualizing fruit quality.

Why it matches plant phenotyping methodsリンゴ果実の乾物含量という植物器官形質を、ハイパースペクトル画像とGA・XAIによる波長選択で予測・可視化する手法が研究の中心である。

abstractthis study presents an inventive method combining a genetic algorithm (GA) with explainable artificial intelligence (XAI) to select key wavelengths for predicting apple dry matter content (DMC).

Measurement of Fruit Diameter Using RGB-D Cameras for the Purpose of Fruit Growth Assessment →

This paper describes a method for measuring the diameter of fruits with a near spherical shape using an RGB-D camera in order to investigate fruit enlargement at different times of the year. In general depth-based measurement methods, when considering the diameter of an object with a near-spherical shape, the diameter can be calculated by obtaining the Euclidean distance from the coordinates of the object's sides or by using the depth at the center of the object and the size of the object on the RGB image. However, it is difficult to accurately determine the diameter of an object because of errors in the calculated results due to the perspective projection of a general camera. Therefore, this study proposes a method to measure the diameter of fruits that have a shape similar to a sphere. Although this study focuses on young apple fruits, the proposed method can be applied to other agricultural crops, as well as to objects that are similar to spheres. In addition, we have also studied a correction that takes into account the rotation of the object so that the method can be applied to objects with circular cross-sections.

Why it matches plant phenotyping methodsRGB-D画像から果実径を推定する手法の開発が研究の中心であり、植物器官の形態形質を直接測定するため。

abstractThis paper describes a method for measuring the diameter of fruits with a near spherical shape using an RGB-D camera

DeepPhenoTree – Apple Edition: a Multi-site apple phenology RGB annotated dataset with deep learning baseline models →

Abstract In machine learning–driven plant phenotyping, well-annotated image datasets are essential for developing robust models capable of capturing phenological variability across environments. Here, we introduce DeepPhenoTree – Apple Edition, a multi-site, multi-variety RGB image dataset dedicated to the detection of key phenological stages in apple trees. The dataset comprises 48,320 time-stamped RGB images acquired across four European orchards of the Apple REFPOP consortium under contrasting climatic conditions. From this large corpus, a carefully curated subset of 808 representative images was manually annotated. It includes 241,600 expert annotations covering developmental stages from dormant bud to fruit maturity. Images were acquired using a standardized tractor-mounted phenotyping platform equipped with active flash illumination, ensuring consistent lighting conditions across sites and acquisition dates. Phenological structures were annotated following the BBCH scale, with bounding boxes adapted to organ visibility and developmental stage. In addition to the dataset, we provide baseline deep learning experiments to illustrate detection performance and assess model generalization across locations.

Why it matches plant phenotyping methodsリンゴの生育ステージを検出する注釈付き画像データセットと、標準化された撮像プラットフォームおよびベースラインモデルを提供しており、植物フェノタイピング手法・再利用可能データが研究の中心である。

abstractHere, we introduce DeepPhenoTree – Apple Edition, a multi-site, multi-variety RGB image dataset dedicated to the detection of key phenological stages in apple trees.

Real-time identification and quantification of apple scab on fruit in preharvest and postharvest conditions using YOLO11: a deep learning approach. →

Background Apple scab (AS), caused by the fungal pathogen Venturia inaequalis, is a major disease of apple that manifests as lesions on leaves and fruits. The disease compromises fruit quality and yield, leading to substantial economic losses. Traditional AS assessment relies on visual scoring, which is labor-intensive, subjective, and poorly reproducible. This study proposes a deep learning-based framework to overcome these limitations and to enable an accurate, scalable AS phenotyping approach. Results Deep learning techniques were employed for the object detection and segmentation of AS symptoms in apple fruits. A two-stage fine-tuning process was applied to color images collected under orchard and laboratory conditions using the YOLO foundation model (YOLO11). The model was first trained to detect healthy apple fruits (Model 1) and subsequently refined to segment AS lesions (Model 2) using high-resolution imagery (864 × 864 pixels). Model 1 (Fruit Detection) achieved 0.98 precision, 0.95 recall, and 0.94 mAP50. Model 2 (Lesion Segmentation) achieved 0.64 precision, 0.75 recall, and 0.75 mAP50. The framework supports real-time processing of images and video. Despite challenges such as variable lighting and symptom heterogeneity, the use of high-resolution training data improved the segmentation accuracy (mAP50-95) of fine-scale lesions by over 50% compared to the previous YOLO architecture. Conclusion These results demonstrate that the proposed deep learning-based approach provides a reliable pipeline for automated AS phenotyping. By improving precision and efficiency in both controlled and field environments, the model enhances apple grading assessments and accelerates breeding efforts to identify AS-resistant genotypes. Furthermore, this work establishes a solid foundation for broader applications in real-time plant disease monitoring and future integration of additional apple diseases.

Why it matches plant phenotyping methodsリンゴ果実上の病斑を画像から検出・セグメント化し、植物病害の程度を自動推定する深層学習手法が研究の中心であるため、植物フェノタイピング手法として含める。

abstractThis study proposes a deep learning-based framework to overcome these limitations and to enable an accurate, scalable AS phenotyping approach.

Enhancing annotations for 5D apple pose estimation through 3D gaussian splatting (3DGS) →

• Novel pipeline simplifying pose annotation • Novel method to quantify the occlusion rate was developed • 99.6% reduction in the amount of manual annotations • Training with an occlusion rate ≤ 95% for the labels lead to the best performance • Improved fruit detection and similar pose estimation as state of the art Automating tasks in orchards is challenging because of the large amount of variation in the environment and occlusions. One of the challenges is apple pose estimation, where key points, such as the calyx, are often occluded. Recently developed pose estimation methods no longer rely on these key points, but still require them for annotations, making annotating challenging and time-consuming. Due to the abovementioned occlusions, there can be conflicting and missing annotations of the same fruit between different images. Novel 3D reconstruction methods can be used to simplify annotating and enlarge datasets. We propose a novel pipeline consisting of 3D Gaussian Splatting to reconstruct an orchard scene, simplified annotations, automated projection of the annotations to images, and the training and evaluation of a pose estimation method. Using our pipeline, 105 manual annotations were required to obtain 28,191 training labels, a reduction of 99.6%. Experimental results indicated that training with labels of fruits that are ≤ 95% occluded resulted in the best performance, with a neutral F1 score of 0.927 on the original images and 0.970 on the rendered images. Adjusting the size of the training dataset had small effects on the model performance in terms of F1 score and pose estimation accuracy. It was found that the least occluded fruits had the best position estimation, which worsened as the fruits became more occluded. It was also found that the tested pose estimation method was unable to correctly learn the orientation estimation of apples.

Why it matches plant phenotyping methods3D Gaussian Splattingによる再構成、アノテーション投影、リンゴの姿勢推定を統合した新規パイプラインが研究の中心であり、果実の位置・向きという植物器官形質を抽出・評価している。

abstractWe propose a novel pipeline consisting of 3D Gaussian Splatting to reconstruct an orchard scene, simplified annotations, automated projection of the annotations to images, and the training and evaluation of a pose estimation method.

SRC-YOLOv8n: a lightweight framework for fine-grained apple leaf disease detection with spatial detail preservation and multi-scale feature enhancement. →

Apple leaf disease detection is crucial for maintaining crop health and ensuring food security, yet current detection methods face significant challenges in balancing accuracy with computational efficiency. Existing lightweight detection models struggle with spatial detail preservation and multi-scale feature representation when processing complex disease symptoms with subtle visual characteristics. This study presents SRC-YOLOv8n, a lightweight framework that integrates spatial detail preservation and multi-scale feature enhancement for fine-grained apple leaf disease detection. The framework incorporates four key innovations: the Spatial Detail Attention C2f (SDA-C2f) module that preserves critical spatial information through Space-to-Depth Convolution and SpatialGroupEnhance mechanisms, the Reparameterized Generalized Feature Pyramid Network (RepGFPN) that optimizes multi-scale feature fusion through training-inference decoupling, the Cross-Level Local Attention Head (CLLAHead) that enables effective cross-scale feature interaction, and the Inner-IoU loss function that improves bounding box regression accuracy. Comprehensive evaluation on the Plant-Pathology-2021-FGVC8 and AppleLeaf9 datasets demonstrates that SRC-YOLOv8n achieves superior performance with 94.1% precision, 92.3% recall, 96.1% mAP50, and 93.2% F1 score while reducing parameters by 16.6%, computational cost by 19.8%, and model size by 17.7% compared to baseline YOLOv8n. The framework provides an effective solution for real-world agricultural monitoring applications requiring both high accuracy and computational efficiency.

Why it matches plant phenotyping methodsリンゴ葉の病徴を画像から検出するための軽量画像解析フレームワークを開発し、複数データセットで性能評価しているため、植物病害状態の表現型取得手法が中心である。

abstractThis study presents SRC-YOLOv8n, a lightweight framework that integrates spatial detail preservation and multi-scale feature enhancement for fine-grained apple leaf disease detection.

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.

An attention-augmented lightweight convolutional framework for fine-grained plant leaf disease classification →

In the recent era, the growth of deep learning is inevitable. Various models such as convolutional neural networks (CNNs) and transformers are used widely in images for high classification accuracy. Since the invention of transformers, researchers have widely used novel approaches using transformers to achieve an impressive accuracy. In spite of this, this paper proposes a novel custom lightweight CNN model called Attentive and Lightweight Network (ALNet). ALNet consists of three major blocks: stem, core, and head. The core part is the novel classifier built as an inspiration from various pre-trained models such as ResNet, SENet (Squeeze and Excitation Network), EfficientNet, SqueezeNet, and ShuffleNet. The main objective is to build a model that has a high classification accuracy while reducing the number of parameters. This reduces the size of the model and hence makes it easy to deploy on cloud platforms and use in edge devices. The model was evaluated using 5-fold cross-validation on three different datasets. The primary dataset was a grapevine dataset with an accuracy of 99.78 percent and 100 percent in multi-class and binary classification respectively. To test the robustness of the model, a multi-class classification using the apple dataset achieved an accuracy of 99.95 percent and a binary classification with the cherry dataset achieved an accuracy of 100 percent. ALNet uses only 0.17 million parameters which is 18 times less parameters than the lightest model (SqueezeNet) and it takes only 14 seconds to train each epoch while pretrained models take 17–31 seconds. ALNet requires only 151.98 MFLOPs with a model size of 677.20 KB, making it approximately 18 times smaller than SqueezeNet. On the whole, ALNet is a highly accurate, lightweight model for plant leaf diseases prediction.

Why it matches plant phenotyping methods葉画像から植物病害状態を分類する軽量CNNを開発・検証しており、病害表現型の画像ベース抽出手法が研究の中心である。

abstractthis paper proposes a novel custom lightweight CNN model called Attentive and Lightweight Network (ALNet).

MDGL-DETR: An Efficient Method for Detecting Apple Leaf Diseases by Integrating Global and Local Features →

Abstract To address challenges such as diverse apple leaf disease phenotypes, high background similarity, and complex natural environments, this study proposes an improved MDGL-DETR model based on RT-DETR, aiming to enhance both the accuracy and efficiency of apple leaf disease detection. First, a Multi-Scale Dilated Asymmetric structure combined with Channel Reduction Attention is designed to strengthen extraction capabilities for multi-scale and global features while reducing computational costs. Second, a Directional Shift Adaptive Context module is proposed, which utilizes shift convolution and SoftPool to dynamically focus on disease regions and suppress redundant background interference. Finally, a Global-Local Collaborative Fusion module is constructed to facilitate efficient interaction between local texture and global semantic information, thereby reinforcing feature representation capabilities. Experimental results indicate that the MDGL-DETR model achieves an mAP50 of 89.09%, an increase of 3.31% over the original RT-DETR, while reducing the computational load by 4.39%. Comprehensive evaluations show that the model outperforms other object detection models. The proposed MDGL-DETR provides a novel solution for the efficient detection of apple leaf diseases.

Why it matches plant phenotyping methodsリンゴ葉の病害領域を画像から検出する深層学習モデルを開発し、検出精度と計算効率を比較評価しているため、植物病害状態のフェノタイピング手法が中心です。

abstractthis study proposes an improved MDGL-DETR model based on RT-DETR, aiming to enhance both the accuracy and efficiency of apple leaf disease detection.

MT-WavYOLO: bridging multi-task learning and 3D frustum fusion for non-destructive robotic harvesting of occluded orchard fruits →

One of the key challenges in orchard robots is accurately localizing occluded fruits in complex environments, especially when the fruit targets are split into multiple isolated regions within images. Traditional single-task network models exhibit limited capability in discerning fragmented targets that belong to the same fruit but are segmented into multiple spatially isolated regions within images. In addition, fruit localization largely relies on high-cost sensors or additional 3-D localization algorithms. To address this issue, we propose a fruit detection and centroid localization method based on a Multi-Task Wavelet-Enhanced YOLO (MT-WavYOLO) to enhance the success rate of robotic operations on occluded fruit targets. Initially, a lightweight semantic segmentation branch was integrated into the YOLOv8 backbone network to precisely segment exposed fruits, while retaining the original object detection branch to fully identify occluded fruits. To address the diminished sensitivity of conventional models to geometric profiles of heavily occluded fruits, a novel feature fusion module, C2f_WTConv, was designed by incorporating wavelet transform convolution, leveraging the multi-frequency robustness of wavelet representations to enhance the model’s feature extraction capabilities under complex orchard occlusions. Subsequently, a 3D frustum-based point cloud processing method was proposed, combining the detection results from MT-WavYOLO with the semantic segmentation masks to accurately localize occluded fruits. MT-WavYOLO demonstrated a 2%, 1.5%, and 2.2% improvement in Precision, Recall, and mAP50, respectively, on our custom-built dataset compared to the latest YOLOv10s model. Semantic segmentation performance, measured by Intersection over Union (IoU) and Accuracy, was improved by 5.2% and 3.8%, respectively, over the state-of-the-art Deeplabv3+ network. Compared to the adapted multi-task network YOLOP, MT-WavYOLO achieved a 3.4% increase in mAP50 and a 2.7% improvement in IoU. In addition, MT-WavYOLO has a compact footprint of 10.2 M parameters and achieves approximately 27 FPS in real-time inference, thereby meeting the requirements of robotic harvesting operations. The proposed localization method was evaluated through 600 fruit localization tests using six different RGB-D cameras in an orchard environment. The average experimental results demonstrated that the centroid localization and radius estimation errors were reduced by 42.5%, 73.7%, 16.17%, and 11.25%, respectively, compared to traditional 3D bounding box methods and our previous approaches. These results indicate that the MT-WavYOLO combined with the frustum-based method significantly enhances the accuracy of apple localization under complex orchard conditions using consumer-grade sensors, providing a strong practical foundation for non-destructive robotic harvesting.

Why it matches plant phenotyping methods果実の検出・3D重心定位という植物器官の形態的状態を、画像分割・深層学習・点群処理で推定する手法を開発し、データセットおよび複数カメラで性能評価しているため、方法が中心的である。

abstractwe propose a fruit detection and centroid localization method based on a Multi-Task Wavelet-Enhanced YOLO (MT-WavYOLO)

Integration of X-Ray CT, Sensor Fusion, and Machine Learning for Advanced Modeling of Preharvest Apple Growth Dynamics. →

Understanding the complex interplay between environmental factors and fruit quality development requires sophisticated analytical approaches linking cellular architecture to environmental conditions. This study introduces a novel application of dual-resolution X-ray computed tomography (CT) for the non-destructive characterization of apple internal tissue architecture in relation to fruit growth, thereby advancing beyond traditional methods that are primarily focused on postharvest analysis. By extracting detailed three-dimensional structural parameters, we reveal tissue porosity and heterogeneity influenced by crop load, maturity timing and canopy position, offering insights into internal quality attributes. Employing correlation analysis, Principal Component Analysis, Canonical Correlation Analysis, and Structural Equation Modeling, we identify temperature as the primary environmental driver, particularly during early developmental stages (45 Days After Full Bloom, DAFB), and uncover nonlinear, hierarchical effects of preharvest environmental factors such as vapor pressure deficit, relative humidity, and light on quality traits. Machine learning models (Multiple Linear Regression, Random Forest, XGBoost) achieve high predictive accuracy (R 2 > 0.99 for Multiple Linear Regression), with temperature as the key predictor. These baseline results represent findings from a single growing season and require validation across multiple seasons and cultivars before operational application. Temporal analysis highlights the importance of early-stage environmental conditions. Integrating structural and environmental data through innovative visualization tools, such as anatomy-based radar charts, facilitates comprehensive interpretation of complex interactions. This multidisciplinary framework enhances predictive precision and provides a baseline methodology to support precision orchard management under typical agricultural variability.

Why it matches plant phenotyping methodsリンゴ果実の内部組織構造をX線CTで非破壊・三次元計測し、構造パラメータを抽出する手法が研究の中心であり、環境データとの統合や機械学習による形質推定も行っているため。

abstractThis study introduces a novel application of dual-resolution X-ray computed tomography (CT) for the non-destructive characterization of apple internal tissue architecture in relation to fruit growth

Fusion Driven Apple leaf Disease Classification using a Siamese Squeeze and Excitation Residual Network with Feature Pyramid Learning →

Abstract The classification of apple leaf diseases is a complex task that requires specialized expertise and is a major contributor to crop losses in agriculture worldwide. The effective earlier diagnosis of automatic apple leaf disease classification model is crucial to maintain the crop's healthy growth. The existing research cannot provide adequate classification results for apple leaf diseases due to the sophisticated classification models. This research proposed the novel Feature pyramid Siamese Squeeze and Excitation Residual neural network (FPS2ENet) model for precise categorization of apple leaf diseases. Initially, the proposed research employed the bilateral guided filter (BGF) which is composed of median and fast guided filter to clean up the apple leaf images. Furthermore, the color, textural and neighboring pixel relationship features are extracted from the apple leaf image with aid of windmill graph-based feature extraction technique. After that, the proposed model is employed with two blocks namely Feature pyramid neural network block (FPNN) for deep feature selection as well as Squeeze and Excitation Residual neural network block for disease classification. Moreover, the Siamese layer and fully connected layer are employed for fusion and classifying process respectively. In experimental analysis, there are four datasets are utilized including Kashmiri Apple Plant Disease Dataset (KAPD), New Plant Diseases Dataset (NPD), Plant Pathology Apple (PPA) dataset, plant village-apple-color (PVAC) dataset. In the performance analysis scenario, the various kinds of analysis are evaluated with different existing methods such as Bi-GRU, Bi-LSTM, DenseNet, and ResNet for accessing the proposed model effectiveness. The numerous effective analyses are performance metric analysis, loss and accuracy curve analysis, ROC curve analysis and confusion matrix analysis. In addition to this, the proposed method can attain 98.81%, 99.43%, 99.88% and 99.61% accuracy in KAPD, NPD, PPA and PVAC dataset correspondingly.

Why it matches plant phenotyping methodsリンゴ葉画像から病害状態を分類する画像ベースの表現型推定手法を提案・評価しており、病害分類モデルの開発が研究の中心である。

abstractThis research proposed the novel Feature pyramid Siamese Squeeze and Excitation Residual neural network (FPS2ENet) model for precise categorization of apple leaf diseases.

The digital orchard: advanced data-driven technologies in apple breeding and genetic modification. →

The apple (Malus × domestica), a globally significant perennial fruit crop, faces immense pressure from climate change, evolving pathogens, and consumer demand for novel traits. Also, remains constrained by slow trait selection despite technological advances. Further, the traditional breeding methods are slow and resource-intensive, hampered by the apple's long juvenile period and high heterozygosity. This systematic literature review (SLR) synthesizes the state of the art in advanced data-driven technologies for accelerating apple breeding and genetic modification. Following the PRISMA-EcoEvo protocol, 47 selected studies were analyzed from databases including Web of Science, Scopus, and PubMed. Our thematic synthesis reveals a paradigm shift towards a "digital breeding" model, characterized by the convergence of three core technological pillars. First, high-throughput phenotyping (HTP), which leverages sensor modalities such as RGB-D, hyperspectral imaging, and LiDAR, is automating the collection of trait data at an unprecedented scale. Second, machine learning (ML) and deep learning (DL) algorithms are being deployed for diverse applications, including cultivar identification with over 96% accuracy, non-destructive quality prediction, and genomic selection, thereby boosting predictive ability for key traits by up to 18%. Third, precise and efficient genome editing, predominantly using Clustered Regularly Interspaced Short Palindromic Repeats (CRISPR)/CRISPR-associated protein 9 (Cas9), is enabling the rapid introduction of desirable traits, such as disease resistance, enhanced shelf life, and improved nutrient uptake. Demonstrated transgene-free editing protocols are accelerating the path to commercialization. We further explore the integration of these pillars through the agricultural internet of things (AIoT) and discuss emerging frontiers, including federated learning for data privacy, explainable AI (XAI) for model transparency, and the implications of recent regulatory frameworks. This review identifies critical research gaps, including the need for standardized open-access datasets and integrated end-to-end system validation. It concludes that the synergistic application of these technologies is poised to revolutionize the speed, precision, and resilience of apple improvement programs worldwide.

Why it matches plant phenotyping methodsリンゴ育種におけるデータ駆動技術の系統的レビューであり、高スループット表現型解析のセンサー技術と技術統合・検証課題を主要に扱っている。

abstractThis systematic literature review (SLR) synthesizes the state of the art in advanced data-driven technologies for accelerating apple breeding and genetic modification.

A Prediction Framework of Apple Orchard Yield with Multispectral Remote Sensing and Ground Features. →

Aiming at the problem that the current traditional apple yield estimation methods rely on manual investigation and do not make full use of multi-source information, this paper proposes an apple orchard yield prediction framework combining multispectral remote sensing features and ground features. The framework is oriented to the demand of yield prediction at different scales. It can not only realize the prediction of apple yield at the district and county scales, but also modify the prediction results of small-scale orchards based on the acquisition of orchard features. The framework consists of three parts, namely, apple orchard planting area extraction, district and county large-scale yield prediction and small-scale orchard yield prediction correction. (1) During apple orchard planting area extraction, the samples of some apple planting areas in the study area were obtained through field investigation, and the orchard and non-orchard areas were classified and discriminated, providing a spatial basis for the collection of subsequent yield prediction-related data. (2) In the large-scale yield prediction of districts and counties, based on the obtained orchard-planting areas, the corresponding multispectral remote sensing features and environmental features were obtained using Google Earth engine platform. In order to avoid the noise interference caused by local pixel differences, the obtained data were median synthesized, and the feature set was constructed by combining the yield and other information. On this basis, the feature set was divided and sent to Apple Orchard Yield Prediction Network (APYieldNet) for training and testing, and the district and county large-scale yield prediction model was obtained. (3) During the part of small-scale orchard yield prediction correction, the optimal model for large-scale yield prediction at the district and county levels is utilized to forecast the yield of the entire planting area and the internal local sampling areas of the small-scale orchard. Within the local sampling areas, the number of fruits is identified through the YOLO-A model, and the actual yield is estimated based on the empirical single fruit weight as a ground feature, which is used to calculate the correction factor. Finally, the proportional correction method is employed to correct the error in the prediction results of the entire small-scale orchard area, thus obtaining a more accurate yield prediction for the small-scale orchard. The experiment showed that (1) the yield prediction model APYieldNet (MAE = 152.68 kg/mu, RMSE = 203.92 kg/mu) proposed in this paper achieved better results than other methods; (2) the proposed YOLO-A model achieves superior detection performance for apple fruits and flowers in complex orchard environments compared to existing methods; (3) in this paper, through the method of proportional correction, the prediction results of APYieldNet for small-scale orchard are closer to the real yield.

Why it matches plant phenotyping methodsマルチスペクトル情報、地上特徴、果実検出を統合してリンゴ収量を推定する技術的フレームワークが研究の中心であり、単なる農業実験のルーチン測定ではない。

abstractthis paper proposes an apple orchard yield prediction framework combining multispectral remote sensing features and ground features

A Machine Vision-Based Online Apple Grading System Toward In-Field Sorting →

Highlights A screw-conveyor-based machine vision system was evaluated for online apple grading. The developed vision pipeline enabled multi-view-based size estimation and comprehensive surface defect inspection. Apple sizing accuracy exceeded 95.9% across conveyor speeds of 1–2 apples s-1 per conveyor lane. The system achieved sample-level grading accuracies of up to 95.4%, demonstrating its potential for further integration and in-field validation. ABSTRACT. Apple quality grading is a critical operation in postharvest handling; however, most existing grading systems are designed for controlled packinghouse environments rather than in-field operation at harvest, which can achieve substantial cost savings for both growers and packers and improve postharvest inventory management. To address the need for in-field apple grading technology, building on our prior work, this study developed a machine vision-based apple grading system toward in-field sorting by integrating a screw-conveyor-based fruit handling mechanism with automated defect inspection and size estimation. The system enables continuous fruit transportation and rotation, allowing multi-view image acquisition for comprehensive surface assessment. The vision module comprises an enclosed image chamber equipped with uniform LED illumination and a top-mounted RGB-D (red-green-blue-depth) camera, ensuring stable and consistent quality of acquired imagery. A computer vision-based pipeline was developed to detect, track, and segment individual apples for surface defect evaluation and sizing. Multi-view images acquired during fruit rotation were fused to achieve full surface coverage. In addition, a geometry-based diameter estimation method integrating stem/calyx-aware boundary localization was introduced to improve fruit sizing robustness and accuracy. Experimental results demonstrate diameter estimation accuracy exceeding 95.9% and maintained sample-level grading accuracies of 95.4%, 94.2%, and 92.3% at conveyor speeds of 1, 1.5, and 2 apples s -1 per lane, respectively. These results demonstrate that the proposed system can support rapid apple grading and provide a practical step toward in-field fruit sorting. Both the dataset and software programs of this study has been made publicly available. Keywords: Apple, In-field grading, Machine vision, Multi-view imaging, Online inspection.

Why it matches plant phenotyping methodsリンゴのサイズ推定と表面欠陥評価を行う画像ベースのオンライン表現型取得・選別システムを開発し、精度検証しているため、植物フェノタイピング手法が中心である。

abstractA computer vision-based pipeline was developed to detect, track, and segment individual apples for surface defect evaluation and sizing.

Realtime multi-RGBD SLAM framework for 3D reconstruction and phenotyping in large-scale apple orchards →

Three-dimensional (3D) reconstructions of orchards offer richer data for digital phenotyping and underpin smart-agriculture applications. However, achieving high-level reconstruction quality and robustness is challenging due to the complex structure of the orchard. This study presents a novel framework that provides centimeter-level 3D realtime reconstructions and phenotyping for apple orchards. A multi-RGBD camera array was adopted, creating a wide overlapping view and robust features. The hybrid odometry front-end and the dual loop-closure strategy ensured low-drift pose estimation. The generated pointcloud was input into the ellipsoid-fitting routine to extract fruit diameter and volume. We validated this framework through reconstruction and phenotypic errors in four rows of an apple orchard with different tree spacings. The root mean square error of the absolute trajectory error in global reconstruction was less than 16 mm. The mean absolute percentage error (MAPE) of the local fiducial distance of approximately 5 m was less than 0.12%. The system was implemented at higher than 12.5 frames per second in an embedded system. The MAPEs of the fruit’s diameter were 2–2.17%, and those of its volume were 5.3–5.6%. Additionally, ablation experiments were carried out on multi-camera and loop-closed elements, and comparisons were made with existing methods to further demonstrate their effectiveness. In conclusion, this research provides an efficient and stable deployable solution for 3D reconstruction of orchards, which is conducive to the development of more advanced and multilayer modern orchard models and promotes the practice of smart agriculture.

Why it matches plant phenotyping methodsリンゴ園向けのマルチRGB-Dによる3D再構成と、点群から果実径・体積を抽出するフェノタイピング手法を開発・検証しており、方法が研究の中心である。

abstractThis study presents a novel framework that provides centimeter-level 3D realtime reconstructions and phenotyping for apple orchards.

Improved Butterfly Optimization and Feature Fusion for Apple Leaf Disease Classification →

In apple leaf disease classification, the health of apple trees plays a crucial role in crop care and productivity. Since leaves are responsible for photosynthesis, they are essential for nutrient transport and overall plant vitality. Therefore, accurate disease identification enables timely intervention and effective disease management. This confirms the optimal yield and crop quality. This research introduces a novel parallel priority feature fusion and improved Butterfly Optimization network for apple leaf disease classification. This method uses PlantVillage, Apple Tree Leaf Disease and PlantPathology Apple as three benchmark datasets. Images undergo preprocessing, in which resizing, normalisation, contrast enhancement using the Top‐Hat operation and data augmentation are used to improve strength and reduce data imbalance. High‐level and complementary deep features by using transfer learning, EfficientNet‐B3 and NasNetLarge are two pre‐trained Convolutional Neural Networks are extracted. These features are fused by the parallel priority strategy, which reduces redundancy and enhances discriminability representation. Then, for optimal feature selection, overfitting is reduced and generalisation is improved using an improved Butterfly Optimization Algorithm with adaptive crossover strategies. The selected features are then classified using Light Gradient‐Boosting Machine, which handles high‐dimensional data efficiently, and more robust classification is ensured. The experimental results demonstrate that the proposed method achieves superior performance, attaining an accuracy of 98.97% and an F1‐score of 97.57% confirming its effectiveness in feature representation and disease discrimination. The proposed method enables early and accurate detection of apple leaf diseases, providing an efficient and reliable solution. This contributes to sustainable agricultural management and enhanced productivity.

Why it matches plant phenotyping methodsリンゴ葉の病害状態を画像から分類する深層学習・特徴融合・特徴選択手法を開発し、複数データセットで性能評価しているため、植物フェノタイピング手法が中心である。

abstractThis research introduces a novel parallel priority feature fusion and improved Butterfly Optimization network for apple leaf disease classification.

Dual-branch feature-enhanced neural network for apple SSC estimation from hyperspectral imaging →

Rapid and accurate assessment of apple soluble solids content (SSC) is essential for breeding research and enhancing marketing efficiency. Hyperspectral data provides rich spectral-spatial information for internal quality assessment, while conventional machine learning methods rely on manual feature engineering and linear assumptions, limiting their ability to capture complex spectral characteristics. Although deep learning techniques offer improved representation learning, existing architectures often fail to model global spectral dependencies and lack customized design for hyperspectral data properties. To address these limitations, we propose a dual-branch feature-enhanced network (DBFENet) for rapid, non-destructive estimation of SSC in apples using hyperspectral imaging. DBFENet integrated reconstruction learning and regression tasks within a complementary framework. The reconstruction branch employs a self-supervised autoencoder network to learn features that preserve essential, generalizable spectral information by accurately reconstructing the original input. The regression branch incorporates a 2D attention mechanism to capture long-range spectral dependencies beyond local patterns. This dual-branch design enables more robust and generalized feature extraction from high-dimensional spectral data. Comprehensive experiments demonstrate that DBFENet significantly outperforms six state-of-the-art methods, including PLSR, RR, SVR, 1D-CNN, MLP, and ResNet18-1D, achieving an Rp of 0.9437 and MSE of 0.2485. The results validate DBFENet as an effective tool for non-destructive SSC evaluation, providing a significant advancement in hyperspectral data analysis for agricultural product quality monitoring.

Why it matches plant phenotyping methodsリンゴ果実のSSCという植物器官形質を対象に、ハイパースペクトル画像から非破壊推定するニューラルネットワークを開発し、既存手法と比較検証しており、表現型取得・推定法が中心である。

abstractwe propose a dual-branch feature-enhanced network (DBFENet) for rapid, non-destructive estimation of SSC in apples using hyperspectral imaging.

Enchancing Apple Plant Leaf Disease Detection Performance with Transfer Learning Methods →

It is very important in agriculture to detect diseases in plants and recovery solutions to produce more crop and to improve efficiency. Enhancements in automated disease detection and analysis can offer significant advantages for taking prompt action, enabling interventions at earlier stages to treat the disease and prevent its spread. This proactive approach could help minimize damage to crop yields. This research is aimed at improving classification performance for apple plant leaf disease detection using transfer learning approaches. The goal is to take necessary precautions for unhealthy apple plants for productive agriculture and healthy food. It discriminates sick apple plants from healthy counterparts by implementing image processing with apple leaf photographs. In this study, traditional machine learning methods are applied for apple plant disease detection task and the classification achievement scores are maximized with transfer learning techniques. The experiments are conducted on a real-world data set including 3164 apple leaf images. As a result, those experiments reveal that transfer learning methods especially EfficientNetB0 has made a significant improvement on classification accuracy for this task. Accuracy and F-score values obtained by transfer learning methods are over 99% which states that they can be considered reliable for plant disease detection tasks.

Why it matches plant phenotyping methodsリンゴ葉画像から病害状態を推定する画像分類手法が研究の中心であり、転移学習手法の比較・性能評価を行っているため、植物フェノタイピング手法研究に該当する。

abstractThis research is aimed at improving classification performance for apple plant leaf disease detection using transfer learning approaches.

Sparse Representation-based Plant Disease Detection using Leaf Image Processing →

Existing image-based crop disease recognition approaches commonly determine multiple feature types from images of diseased plant leaves. Yet, a shared limitation remains: those discriminative features are commonly assumed to contribute uniformly to the classification decision. In this study, we present an apple leaf disease recognition pipeline composed of three sequential modules: segmenting diseased leaf images through K-means clustering, deriving shape and color features from the diseased region, and then performing classification of diseased leaf images via the sparse representation (SR) method. One notable strength of this pipeline is that conducting classification within the SR domain can notably decrease computational load while simultaneously boosting recognition accuracy. We benchmark this method against three baseline classifiers - support vector machines (SVM), artificial neural networks (ANN), and decision tree (DT) - using a leaf image dataset containing three apple leaf disease categories: black spot, frogeye leaf spot, and cedar apple rust. Experimental results indicate that the presented pipeline attains 93.9% overall accuracy, whereas the SVM, ANN, and DT achieve 76.8%, 83.7%, and 82.9%, respectively.

Why it matches plant phenotyping methodsリンゴ葉の病斑領域を画像から分割・特徴抽出し、病害状態を分類する画像ベースの植物フェノタイピング手法を開発・比較しており、方法が中心である。

abstractwe present an apple leaf disease recognition pipeline composed of three sequential modules: segmenting diseased leaf images through K-means clustering, deriving shape and color features from the diseased region, and then performing classification of diseased leaf images via the sparse representation (SR) method.

A Precise Apple Quality Prediction Model Integrating Driving Factor Screening and BP Neural Network. →

Apple fruit quality is primarily determined by Vitamin C (VC), Soluble Saccharides (SSs), Titratable Acid (TA), and the Soluble Saccharides/Titratable Acid (SSs/TA). This study aims to establish a prediction model based on the Back Propagation (BP) neural network by analyzing the intrinsic relationships between these quality indicators and the photosynthetic physiological characteristics of fruit trees, providing a new method for the precise prediction and regulation of fruit quality. Using 'Fuji' apple as the material, fruit quality indicators, leaf photosynthetic parameters, canopy structure indicators, and carbon-water-nitrogen metabolism indicators were systematically measured. Correlation analysis was employed to identify key influencing factors, BP neural network models with different hidden layer structures were constructed, and the optimal feature subset was screened through feature importance analysis, single-factor sensitivity analysis, and ablation experiments, ultimately establishing a simplified and efficient prediction model. Pn, Gs, SPCI, and DUE showed significant positive correlations with VC, SS, and SS/TA, whereas N and NLT were significantly positively correlated with TA content. SUE was identified as a common core driving factor for VC, SS, and SS/TA. The BP neural network demonstrated strong predictive performance for the four quality indicators, with the optimal model achieving validation set R 2 values of 0.87, 0.86, 0.86, and 0.89, respectively. The simplified model developed through feature screening exhibited further improved performance: the validation set R 2 for the VC prediction model increased to 0.93, while MAE and MAPE decreased by 32% and 35%, respectively. Photosynthetic characteristics and nitrogen metabolism status of the fruit trees serve as key physiological foundations determining apple quality. The quality prediction model based on the BP neural network achieved high accuracy, and its predictive performance was significantly enhanced after feature refinement, providing an effective tool for precise apple quality prediction and smart orchard management.

Why it matches plant phenotyping methodsリンゴ果実品質という植物形質を対象に、BPニューラルネットワーク、特徴量選択、感度分析、アブレーション実験を組み合わせた予測手法を開発・検証しており、形質推定法が研究の中心である。

abstractThis study aims to establish a prediction model based on the Back Propagation (BP) neural network

Multimodal learning on RGB-D image for precise litchi phenotyping and weight estimation →

Accurate measurement of key phenotypic traits, including the horizontal and vertical diameters, the weights of both fruit and pit, is essential for the selection of elite litchi cultivars and the advancement of breeding research. Manual measurement, however, is laborious, inefficient, and subjective, highlighting the urgent need for automated and precise phenotyping tools. Unlike apples, mangoes, and grapes, litchi combines a spiny, highly variable pericarp (heterogeneous areoles/tubercles across cultivars) with diverse seed morphology (including irregular, wrinkled aborted seeds), thereby increasing the difficulty of semantic segmentation and biasing diameters and weight estimation. This study presents LitchiPhenoNet, a multimodal learning framework for litchi phenotypic analysis that employs a dual-branch architecture integrating RGB (color/texture) and depth (spatial/structural) information. Experiments were conducted on an RGB-D dataset comprising 1,198 image pairs (1280×720) across 10 cultivars, using a stratified train/test split of 958/240 pairs by cultivar. To address inherent semantic and scale inconsistencies between modalities, the framework incorporates the RD-Fusion module for precise cross-modal feature extraction, improving robustness under complex and variable pericarp surfaces. Comparative experiments show that LitchiPhenoNet consistently outperforms leading YOLO-based models, achieving millimeter-level diameter estimation with coefficients of determination approaching 0.98 and mean errors within 2 mm. For weight estimation, gram-level precision is attained across whole fruit, pit, and pulp, with coefficients of determination up to 0.98 and mean errors comparable to repeated manual measurements. By handling fine-scale surface relief and cross-cultivar variability, the framework is readily extensible to other textured fruits and scalable for high-throughput phenotyping in breeding programs. Collectively, these results demonstrate that LitchiPhenoNet provides an efficient, reliable, and accurate solution for quantifying litchi phenotypic traits, substantially advancing the objectivity and efficiency of phenotypic analysis and breeding selection.

Why it matches plant phenotyping methodsRGB-D画像を用いてライチ果実・種子・果肉の径と重量を自動推定する専用フレームワークを開発し、複数品種・比較実験で性能検証しているため、植物表現型取得法が中心である。

abstractThis study presents LitchiPhenoNet, a multimodal learning framework for litchi phenotypic analysis that employs a dual-branch architecture integrating RGB (color/texture) and depth (spatial/structural) information.

Early detection of apple bruises using spectral-spatial enhanced 3D CNN and region-based hyperspectral analysis →

Hyperspectral imaging (HSI) has revolutionized the non-destructive detection of fruit defects by integrating spectral and spatial information. However, detecting early-stage minor bruises in apples remains challenging due to weak hyperspectral signals, high data dimensionality, and low computational efficiency. To address these issues, this study proposes a novel hyperspectral apple bruise detection network, termed 3D-HDI. The model constructs a backbone network using multiple 3D convolutions and integrates a feature enhancement module with a path aggregation network to amplify spectral signals and improve damage differentiation. Furthermore, it replaces pixel-by-pixel classification with a region-based detection head, significantly enhancing computational efficiency and accuracy. Experimental results demonstrate that the proposed model achieves a higher recognition rate (96.25%) while maintaining comparable detection efficiency, outperforming traditional classification networks such as 3D-EfficientNet, 3D-MobileNet, and 3D-AlexNet. This research advances the application of HSI and deep learning in fruit quality assessment, providing a robust solution for early-stage bruises detection.

Why it matches plant phenotyping methodsリンゴ果実の打撲という植物器官の状態を、ハイパースペクトル画像と新規3D CNN・領域検出手法で推定する方法開発が中心である。

abstractTo address these issues, this study proposes a novel hyperspectral apple bruise detection network, termed 3D-HDI.

Methodology for the assessment of leaf area in fruit tree orchards using a terrestrial LiDAR-based system →

Accurate estimation of canopy geometric and structural characteristics, such as leaf area (LA), is essential for improving resource efficiency in fruit tree crop management. LA is a key biophysical parameter, influencing physiological processes like carbon fixation, evapotranspiration, and light interception, as well as fruit quality and yield. However, its measurement is complex due to the substantial number of leaves and the three-dimensional nature of tree canopies.An alternative approach, the Projected Tree Row Surface (PTRS), has shown a strong correlation with LA and has been recognized by the scientific community. Despite its robustness, the original PTRS method requires time-consuming manual data collection, which limits its practical application in the field.This study introduces a novel automated methodology for calculating the PTRS, validated using high-resolution ground-truth data providing LA values at 0.1-m intervals along the tree rows. When evaluated on almond, pear, and apple trees as well as vineyards, the method achieved remarkably high correlations between PTRS and LA, with coefficients up to r = 0.97 and r = 0.99 at optimal resolutions (0.1 –0.2 m PTRS per 1 m row section). These results demonstrate that the approach delivers consistent and reliable measurements of LA under diverse field conditions, enabling real-time, high-resolution assessment of tree-row canopies.The automated PTRSₙ approach enables fast and efficient LA estimation and can be adapted to any point cloud dataset. It supports flexible resolution to balance accuracy and processing time and can be applied to full rows, individual trees, or canopy segments. This methodology represents a step forward in automating LA assessment and supports the development of real-time applications in precision agriculture.

Why it matches plant phenotyping methods果樹・ブドウ樹冠の葉面積を推定するLiDARベースの自動PTRS手法を開発し、実測LAで検証しており、植物形質取得法が研究の中心である。

abstractThis study introduces a novel automated methodology for calculating the PTRS, validated using high-resolution ground-truth data providing LA values at 0.1-m intervals along the tree rows.

A UNet-GAN two-stage network for rapid and accurate prediction of apple optical properties from single multi-frequency images →

Spatial frequency domain imaging (SFDI) is a non-invasive optical imaging technique widely used for the quantitative determination of fruit tissue optical properties, specifically absorption coefficient (μₐ) and reduced scattering coefficient (μₛ’). However, traditional SFDI methods rely on multiple frequency and phase images, limiting real-time imaging capabilities. To address this issue, we present a novel rapid prediction method based on a two-stage deep neural network architecture, termed FSGOP (Frequency-Spatial Attention UNet and GAN-based two-stage network for optical properties prediction). Compared with conventional three-phase demodulation SFDI, this method reduces the acquisition time by approximately 5/6 and requires only 0.21 s for inference. In the first stage, a UNet network enhanced by Frequency-Spatial Attention (FSA) is employed to effectively decouple the multi-frequency components. In the second stage, a Generative Adversarial Network (GAN) is utilized to predict the optical properties, thereby enabling the simultaneous extraction of μₐ and μₛ’ maps under different frequency conditions from a single multi-frequency mixed fringe image. In experiments on apples, pears, and peaches, the method yielded normalized mean absolute errors of 0.10 (f₁) and 0.09 (f₂) for μₛ’, and 0.07 and 0.06 for μₐ, respectively. The results revealed significant complementary information in the optical property maps at different frequencies, with lower frequencies being more sensitive to subsurface damage and higher frequencies revealing surface texture features more effectively. This method enhances information utilization and real-time performance in multi-frequency imaging, offering a rapid, accurate, and low-cost solution for optical property extraction and quality inspection of agricultural products.

Why it matches plant phenotyping methods果実の光学特性を単一画像から推定する画像・深層学習手法を開発し、取得時間と精度を評価しており、植物器官の状態計測が中心である。

abstractwe present a novel rapid prediction method based on a two-stage deep neural network architecture, termed FSGOP

Apple diameter prediction during mechanical picking based on flexible force-sensing and CNN-BiLSTM-Attention method →

Fruit diameter grading is essential for commercialization, packaging, and market sales, directly impacting the value and competitiveness of fruits. Traditional diameter grading is typically performed after harvesting, relying on manual or mechanical methods. This process introduces extra steps and increases the risk of fruit damage during transit, which can reduce economic efficiency. To address this issue, this study introduces a real-time apple diameter grading method utilizing a flexible force-sensing gripper and the CNN-BiLSTM-Attention deep learning network, enabling synchronized intelligent identification and grading of apple diameter during robotic mechanical picking. First, ionogel-based triboelectric nanogenerators (IG-TENG) were developed and mounted on the surface of a three-finger Fin-Ray flexible picking end effector. A contact force monitoring system was established using a modular apple model, and the force sensor was calibrated. This setup allowed for accurate measurement of the contact force between the fingers and the apple. Using a multi-layer perceptron (MLP) to integrate mechanical response data, robotic hand motor stroke, and apple posture information, an apple contact force model was created to accurately predict the actual gripping force under various grasping conditions. Finally, a CNN-BiLSTM-Attention diameter prediction model was designed to deliver real-time, precise fruit diameter estimates. Orchard experiments demonstrated that the apple diameter grading method, combining force sensing with deep learning, achieved a mean absolute error (MAE) of 2.13 mm and a grading accuracy of 92 %, supporting non-destructive gripping and accurate grading. This research addresses the limitations of traditional diameter grading methods, streamlines harvesting steps, enhances efficiency, and offers a cost-effective and reliable solution for non-destructive fruit diameter grading.

Why it matches plant phenotyping methodsリンゴ径という植物器官形質を、力覚センサーと深層学習でリアルタイム推定・等級化する手法の開発、校正、検証が研究の中心である。

abstractFinally, a CNN-BiLSTM-Attention diameter prediction model was designed to deliver real-time, precise fruit diameter estimates.

Curvature sensing strategy for flexible gripper fingers during grasping deformation: enabling in-orchard online apple size grading →

In-orchard apple size grading remains challenging under occlusions, variable illumination, and irregular fruit morphology. We present a contact–curvature strategy that integrates bending sensors into fin–ray flexible fingers to estimate fruit size during grasping, unifying grasp–and–grade operations. Neural Radiance Fields (NeRF) reconstruction provides offline ground-truth curvature for calibration, yielding strong linear agreement between sensor readings and true curvature (R2=0.9852). Using temporal curvature features across approach, grasp, and steady phases, a gradient-boosting regression model predicts fruit diameter with R2=0.9577 and RMSE = 1.19 mm on the test set. In laboratory conditions, the system achieved an overall grading accuracy of 98.0 % for 200 apples classified into four grades, with a processing capacity of approximately 6 apples·min⁻¹, meeting real-time requirements. In a small-scale orchard pilot study, the system maintainedR2=0.94 andRMSE=1.27 mm, achieving 96 % grading accuracy versus 77 % for a camera-only approach. Compared with vision-only sizing methods, contact-curvature sensing demonstrates inherent robustness to occlusion and illumination while better tolerating morphological irregularities. A methylene–blue protocol confirmed non–destructive operation. Contact–curvature sensing is robust to occlusions/illumination and can, in principle, extend to other near–spherical crops.

Why it matches plant phenotyping methods果実径という植物器官形質を、接触・曲率センサーと回帰モデルで推定する手法を開発し、校正・精度検証・圃場評価まで行っており、表現型取得法が研究の中心である。

abstractWe present a contact–curvature strategy that integrates bending sensors into fin–ray flexible fingers to estimate fruit size during grasping, unifying grasp–and–grade operations.

Prediction of Apple Quality Indicators Under Different Bagging Treatments Using Hyperspectral Imaging Integrated With a Stacking SDAE-PLSR-RR Deep Learning Model. →

The color indices (L*, a*, and b*) and soluble solids content (SSC) serve as essential quality indicators for apples, yet conventional destructive detection methods lack the efficiency required for rapid sorting of apples with varied bagging treatments. To address this limitation, this study proposes a novel stacking model, termed SDAE-PLSR-RR, which integrates hyperspectral imaging with deep learning. Hyperspectral imaging comprehensively captured spectral-spatial features from 307 Fuji apples subjected to three bagging treatments (non-bagged, mesh-bagged, and paper-bagged), enabling systematic analysis of quality-related characteristics. The SDAE-PLSR-RR employs a stacked structure where two parallel, base-level expert models capture complementary features: one Partial Least Squares Regression (PLSR) model processes linear trends in original wavelengths data, while the other analyzes non-linear deep features from a Stacked Denoising Autoencoder (SDAE). A top-level Ridge Regression (RR) model then acts as a meta-learner to fuse the predictions from these two base models, generating a final, more robust output. The integrated SDAE-PLSR-RR model achieved enhanced prediction accuracy for all quality indicators (R 2 p > 0.84), outperforming full-spectrum (R 2 p > 0.73) and feature-wavelength-based models (R 2 p > 0.75). The experimental findings validated the applicability and efficacy of integrating hyperspectral imaging systems with neural network models for non-destructive detection of the quality indicators of apples with different bagging treatments.

Why it matches plant phenotyping methodsリンゴの色指標とSSCという植物器官形質を対象に、ハイパースペクトル画像と新規スタッキングモデルによる非破壊推定法を開発・検証しており、フェノタイピング手法が中心である。

abstractthis study proposes a novel stacking model, termed SDAE-PLSR-RR, which integrates hyperspectral imaging with deep learning.

Physiology at work to model apple expansion growth and skin pigment changes →

Fruit sizing is a major factor in determining yield while non-destructive monitoring of the fruit skin colour and internal quality attributes on the tree can provide valuable maturity and quality information for precision horticulture. Repeated spectral scanning and fruit sizing data of ‘Braeburn’ apples were collected on the tree from about 60 days after flowering until harvest. Assessed variables were: fruit diameter, dry matter (DMC), soluble solids content (SSC), a normalised difference vegetation index (NDVI) and a normalised anthocyanin index (NAI) analysed using indexed non-linear regression based on an adapted von Bertalanffy model (diameter, DMC, SSC), or a logistic model (NDVI, NAI). The reaction rate constants in the models were estimated in common for all fruit in a selection, while the biological shift factors (Δt) estimated the development stage or fruit maturity per individual fruit. Explained parts (R²ₐdⱼ) range from 85 to 97%. Tree location or crop load treatment only minimally affected the rate constants but did affect the estimated Δt values that describe almost all variation in the data. There is a close relationship between the Δt values for diameter, DMC and SSC but less with those of NDVI and almost none with the NAI. These data support the assumption that there is only one stage of fruit maturity, but it is estimated slightly differently depending on the measured variable. The actual relative growth rate strongly depends on the current size. Understanding apple expansion growth will therefore require a closer focus on the cell production period.

Why it matches plant phenotyping methodsリンゴ果実の非破壊スペクトル測定・果径測定と回帰モデルを組み合わせ、果実の成長・成熟・色素変化を個体ごとに推定する技術的ワークフローが中心であるため、植物フェノタイピング手法の応用として含める。

abstractnon-destructive monitoring of the fruit skin colour and internal quality attributes on the tree can provide valuable maturity and quality information for precision horticulture.

Design of a Portable Nondestructive Instrument for Apple Watercore Grade Classification Based on 1DQCNN and Vis/NIR Spectroscopy. →

To address the challenge of nondestructively identifying watercore disease in apples during growth and maturation, a portable device was developed for real-time grading of apple watercore using visible/near-infrared (Vis/NIR) spectroscopy combined with a one-dimensional quadratic convolutional neural network (1DQCNN). The instrument enables rapid, nondestructive, and accurate detection of apple watercore grades. The AI-OX2000-13 micro-spectrometer is used as the core data acquisition unit, and an ARM processing system is built with the STM32F103VET6 as the main control chip. A 4G wireless communication module enables efficient and stable data transmission between the processor and computer, meeting the real-time detection needs of apple watercore content in orchard environments. To improve the scientific and accurate classification of watercore grades, this paper combines the BiSeNet and RIFE algorithms to construct a 3D model of apple watercore, allowing quantification of the degree of watercore and classification into four levels. Based on this, quadratic convolution operations are incorporated into a one-dimensional convolutional neural network (1DCNN), leading to the development of the 1D quadratic convolutional neural network (1DQCNN) model for watercore grade classification. Experimental results indicate that the model achieves a classification accuracy of 98.05%, outperforming traditional methods and conventional CNN models. The designed portable instrument demonstrates excellent accuracy and practicality in real-world applications.

Why it matches plant phenotyping methodsリンゴの水心症状の程度を可搬型Vis/NIR装置と画像・深層学習で定量・分類する計測手法および装置の開発が研究の中心であり、植物病害状態の表現型取得に該当する。

abstracta portable device was developed for real-time grading of apple watercore using visible/near-infrared (Vis/NIR) spectroscopy combined with a one-dimensional quadratic convolutional neural network (1DQCNN).

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.

Utilizing an Efficient Deep Convolutional Neural Network for the Automatic Classification of Plant Leaf Diseases →

Globally, India is second in terms of tomato and apple production. On an average of nearly, 18.399 tons of tomatoes and 30.500 tons of apples are produced in India. Likewise, after rice and wheat, maize scores third most significant food in India and it is one of the top crop producers in the world. The growth and health of these plants are typically afflicted by the diseases. Variety of leaf diseases available under apple, tomato and maize leaves which affect the production. This paper proposes an efficient deep convolutional neural network for identifying and detecting diseases in plant leaves through automatic image classification. The proposed model’s primary goal is to pinpoint a fix for the issue with leaf diseases that affect maize, tomatoes, and apples. The proposed efficient model comprises of residual layers, modified residual layers and global average pooling layers that yield better efficiency in terms of feature extraction and high throughput. The performance of the proposed model is evaluated through training and testing using the Plant Village dataset, which was sourced from a GitHub repository. In addition, the proposed model is evaluated through standard classification metrics and achieved better accuracy in the range 97% to 99%.

Why it matches plant phenotyping methods植物葉の病害状態を画像から自動分類する深層学習手法が研究の中心であり、病害表現型の取得・推定に該当する。

abstractThis paper proposes an efficient deep convolutional neural network for identifying and detecting diseases in plant leaves through automatic image classification.

SAM-Based Leaf Segmentation with Morphological Quality Assessment for Enhanced Plant Disease Detection →

Plant diseases threaten global food security, yet traditional visual inspection often fails to detect early-stage symptoms critical for timely intervention. While deep learning models have shown promise for automated disease detection, their performance often degrades in realistic field conditions. This study investigates whether data-centric preprocessing improves apple leaf disease detection. We present the first systematic evaluation of the Segment Anything Model (SAM) combined with morphological quality assessment for leaf segmentation, compared against whole-image classification using the Plant-Pathology FGVC7 dataset (3,642 apple orchard images). To ensure segmentation reliability, we introduce a five-metric morphological framework (area ratio, aspect ratio, spatial coverage, centroid proximity, border penalty). Experiments with ResNet-18 under 3-fold cross-validation reveal class-specific effects: SAM improves F1 by 3.0% for the minority multiple diseases class, but decreases by$1. 4 {\%}$for healthy leaves where contextual cues aid detection. Rust and scab remain stable above 95% F1, reflecting their distinctive visual signatures. GradCAM ++ confirms that preprocessing redirects attention toward diseaserelevant regions, particularly in complex multiple-disease cases. Overall, these findings show that adaptive preprocessing, rather than universal background removal, offers practical benefits for precision agriculture.

Why it matches plant phenotyping methodsSAMによる葉画像セグメンテーションと形態学的品質評価を中心に、植物病害検出への有効性を体系的に検証しているため、植物表現型取得・抽出手法として収録する。

abstractWe present the first systematic evaluation of the Segment Anything Model (SAM) combined with morphological quality assessment for leaf segmentation

An enhanced deep learning-based framework for diagnosing apple leaf diseases. →

Timely and correct identification of diseases in the apple leaf is also important in protecting crop production and sustaining agriculture. This paper introduces E-YOLOv8, a lightweight improved version of YOLOv8, that can be implemented in real-time and with a limited resource base. The model has three key contributions: (1) GhostConv and C3 fusion to reduce redundant feature extraction and computational cost, (2) CBAM attention and a specifically designed FPN to maximize multi-scale feature fusion and small-lesion detections, and (3) large-scale evaluation on datasets of apple leaf disease, as well as ablation experiments and operational testing on edge devices to verify the accuracy and viability of this model. In experiments, E-YOLOv8 reaches 93.9mAP0.5 using 5.3 GFLOPs and 1.8 M parameters, a 33.9x factor smaller than that of YOLOv8l. These results indicate that E-YOLOv8 has achieved better performance than recent state-of-the-art detectors and is still applicable to practical real-world agricultural tasks.

Why it matches plant phenotyping methodsリンゴ葉の病斑を画像から検出する深層学習手法を開発し、データセット評価、アブレーション、エッジデバイス試験で検証しているため、植物病害状態の表現型取得が中心である。

abstractThis paper introduces E-YOLOv8, a lightweight improved version of YOLOv8, that can be implemented in real-time and with a limited resource base.

Nondestructive Detection of Soluble Solids Content in Apples Based on Multi-Attention Convolutional Neural Network and Hyperspectral Imaging Technology. →

Soluble solids content is the most important attribute related to the quality and price of apples. The objective of this study was to detect the soluble solids content (SSC) in 'Fuji' apples using hyperspectral imaging combined with a deep learning algorithm. The hyperspectral images of 570 apple samples were obtained and the whole region of apple sample hyperspectral data was collected and preprocessed. In addition, a method involving multi-attention convolutional neural network (MA-CNN) is proposed, which extracts spectral and spatial features from hyperspectral images by embedding channel attention (CA) and spatial attention (SA) modules in a convolutional neural network. The CA and SA modules help the network adaptively focus on important spectral-spatial features while reducing the interference of redundant information. Additionally, the Bayesian optimization algorithm (BOA) is used for model hyperparameter optimization. A comprehensive evaluation is conducted by comparing the proposed model with CA-CNN models, SA-CNN, and the current mainstream models. Furthermore, the best prediction performances for detecting SSC in apple samples were obtained from the MA-CNN model, with an Rp2 value of 0.9602 and an RMSEP value of 0.0612 °Brix. The results of this study indicated that the MA-CNN algorithm combined with hyperspectral imaging technology can be used as an effective method for rapid detection of apple quality parameters.

Why it matches plant phenotyping methodsリンゴの可溶性固形分という果実形質を、ハイパースペクトル画像と深層学習で非破壊推定する手法を開発・比較評価しており、形質取得法が中心である。

abstractThe objective of this study was to detect the soluble solids content (SSC) in 'Fuji' apples using hyperspectral imaging combined with a deep learning algorithm.

Plant Disease Detection Using Machine Learning: A Comprehensive Framework and Performance Analysis →

Abstract - The global agricultural sector faces significant challenges due to plant diseases that threaten food security and sustainable agriculture. Traditional methods of disease detection are often labour-intensive, time-consuming, and require specialized expertise. This research presents a comprehensive machine learning framework for automated plant disease detection, leveraging both traditional machine learning and deep learning approaches. We implemented and evaluated multiple models including VGG19, Inception v3, Support Vector Machines (SVM), and k-Nearest Neighbors (kNN) on four distinct datasets: Banana Leaf, Custard Apple Leaf and Fruit, Fig Leaf, and Potato Leaf. Our experimental results demonstrate remarkable performance variations across different crops, with the highest achievement of 99.1% accuracy using VGG19 with kNN on the Custard Apple dataset, while the Potato Leaf dataset presented the greatest challenges with 62.6% accuracy using Inception v3 with SVM. The study provides valuable insights into model selection for specific agricultural applications and highlights the importance of customized solutions based on crop-specific characteristics. We also address critical challenges including dataset limitations, computational requirements, and implementation barriers in real-world agricultural settings. Keywords - Plant disease detection, machine learning, deep learning, convolutional neural networks, agricultural technology, precision agriculture.

Why it matches plant phenotyping methods植物病害状態を対象に、機械学習・深層学習による自動検出フレームワークを提示し、複数モデルとデータセットで性能評価しているため、病害表現型の取得・判定手法が中心である。

abstractThis research presents a comprehensive machine learning framework for automated plant disease detection, leveraging both traditional machine learning and deep learning approaches.

Deep learning technique for plant disease classification and pest detection and model explainability elevating agricultural sustainability. →

The rapid advancement of technologies such as artificial intelligence (AI), deep learning, and precision agriculture tools is driving the development of efficient, data-driven crop management solutions. These innovations are increasingly critical in modern agriculture, where early and accurate detection of plant diseases plays a vital role in securing crop yields and sustainability. Agronomists, agriculturists, and local farmers continue to face significant economic losses due to delayed diagnosis or misclassification of diseases affecting high-value crops, key contributors to the global market. Failure to identify and manage such diseases in time can severely impact both agricultural productivity and global food supply chains. To achieve the United Nations’ sustainable development goals of zero hunger, climate change, good health, and well-being, early and timely disease detection is critical to ensure increased apple-related production, damage control, and reduced application of inappropriate herbicides that pollute the environment. Despite the availability of various methods for early disease detection and classification, how early signs of green attacks can be identified remains uncertain. Using the Turkey Plant Pests and Diseases (TPPD) dataset with 4,447 images categorized into 15 diverse classes, this research implements ResNet-9 to detect and classify the commonly known pests and diseases of six plants, including Malus pumila, Prunus armeniaca, Prunus padus, Prunus persica L. Batsch., Pyrus communis L., and Juglans regia. A laborious hyperparameter tuning, hyperparameter optimization, and augmentation procedure on the training set was done for some imbalanced dataset classes. Testing results of the proposed model demonstrated accuracy, precision, recall, and F1-score values of 97.4%, 96.4%, 97.09%, And 95.7%, respectively, which is a significant leap in comparison to other existing research. This study further elucidates and enhances the interpretability of the proposed model by making saliency maps available using SHapley Additive exPlanations (SHAP) that efficiently illustrate the rationale behind the model’s prediction capabilities. The study further tested the statistical significance of the model, the Area Under the receiver operating characteristic curve (AUC-ROC), and the confidence interval (CI). Critical observations revealed that the model uses several visual cues for disease detection and classification, including (i) edge contours and shape structures that help define lesion boundaries, (ii) texture and color variations that signal symptom type and severity, and (iii) high-activation regions that indicate areas of strong feature relevance. These cues collectively guide the model in distinguishing between visually similar disease patterns across different plant parts. The application of SHAP saliency maps further enabled interpretation by visually localizing and quantifying the influence of these features on the model’s predictions.

Why it matches plant phenotyping methods植物の病害・害虫状態を画像から分類する深層学習手法を開発・評価しており、病徴の局在化と重症度に関連する視覚特徴も解析しているため、植物フェノタイピング手法が中心である。

abstractthis research implements ResNet-9 to detect and classify the commonly known pests and diseases of six plants

The impacts of tree shape, disease distribution and observation geometry on the performances of disease spectral indices of apple trees →

Vegetation indices (VIs) are widely employed in remote sensing for quantitative monitoring of plant disease due to their simplicity and robustness. However, factors such as canopy structure, the distribution of diseases in the canopy, and observation geometry may influence the spectral response of diseased canopies, potentially affecting the performance of VIs developed under specific disease conditions (e.g., early-stage). To date, fewer comprehensive analytical strategy has been proposed to quantitatively assess the confounding effects of multiple factors, which has hindered the selection of optimal VIs for practical disease monitoring. This study proposes an integrated analytical strategy that combines a three-dimensional radiative transfer model (3D RTM) with a multi-criteria decision-making method — entropy-weighted Technique for Order Preference by Similarity to Ideal Solution (TOPSIS) — to systematically evaluate existing disease-related VIs at the canopy scale, based on simulation outputs and ground measurements. We employed the LargE-Scale remote sensing data and image Simulation framework (LESS) to simulate the bidirectional reflectance factor (BRF) of canopies affected by two representative apple diseases, quantitatively evaluated the confounding effects of tree shape, disease distribution and observation geometry on VIs, and systematically ranked the performance of 40 VIs from two critical perspectives. Results from analyses on two disease types showed that Health Index 2014 (HI2014) and Water Band Index in SWIR (WBISWIR) were the top-performing indices for monitoring apple blotch disease (AMB) and apple mosaic disease (MD), respectively, Notably, WBISWIR emerged as the co-optimal index, exhibiting the highest monitoring efficacy across both diseases. Among all indices, the Normalized PRI (PRIn) demonstrated the greatest robustness against variations in tree shapes and disease distributions. WBISWIR exhibited good performance across diverse observation geometries. When comparing the relative influence of three factors on VI performance, tree shape and disease distribution exerted greater effects than observation geometry. Our findings highlight the complex interactions between VIs and confounding factors, emphasizing the necessity of caution when applying disease-related VIs and advocate for comprehensive consideration of tree shape and stress distribution effects during VI selection, especially for early-stage disease detection. This study offers a robust methodological framework for selecting VIs tailored to specific disease and vegetation characteristics, enhancing the precision of remote sensing-based plant disease assessments.

Why it matches plant phenotyping methodsリンゴ樹の病害状態を対象に、3D放射伝達モデル、シミュレーション、実測データ、TOPSISを統合して病害スペクトル指標を評価・選定する方法論が中心であり、植物病害フェノタイピング手法に該当する。

abstractThis study proposes an integrated analytical strategy that combines a three-dimensional radiative transfer model (3D RTM) with a multi-criteria decision-making method — entropy-weighted Technique for Order Preference by Similarity to Ideal Solution (TOPSIS) — to systematically evaluate existing disease-related VIs at the canopy scale, based on simulation outputs and ground measurements.

An Integrated Deep Learning and Sparse Representation Framework for Apple Leaf Disease Recognition →

Proper identification of apple leaf diseases is an important issue towards safeguarding crop production as well as management sustainability of orchards. Manual diagnosis is widely used, but not efficient, subjective, and cannot be easily used to diagnose large-scale monitoring. This paper presents a Hybrid Deep Learning and Sparse Classification Framework, which combines the use of YOLOv8 to extract deep features with L1-regularized machine learning models to do the efficient and interpretable disease recognition. A dataset of 4,006 high-resolution images of apple leaf which had a total of 9 classes which included Alternaria leaf spot, Brown spot, Frogeye leaf spot, gray spot, Healthy, Mosaic, Powdery mildew, Rust, and Scab, was used. After 60 epochs, YOLOv8 attained a mean Average Precision (MAP) of 98.10% at 0.5, a precision value of 95.16%, and a recall of 95.97, which is quite good in terms of detection. The 3,648 resulting dimension feature space was then narrowed down to 319 dimensions using Lasso feature selection and 13.48 percent of sparsity and without losing important discriminatory data. The Lasso model achieved the best test accuracy of 97.50 percent and highest cross-validation was 98.78 percent, which is better than the Logistic Regression (L1) and random forest baselines among the compared classifiers. Sparse regularization integration did not only reduce overfitting but also led to an improved model interpretability, by forcing the visual patterns to focus on diseases. The suggested model is a computationally efficient, scalable, and interpretable AI model to real-time agricultural diagnostics. It offers a bright future to the precision farming systems that can not only detect diseases early but also protect their yields and make better decisions based on interpretable hybrid intelligence.

Why it matches plant phenotyping methodsリンゴ葉の病徴を画像から認識する深層学習・特徴選択フレームワークが研究の中心であり、植物病害状態の画像ベース表現型推定に該当する。

abstractThis paper presents a Hybrid Deep Learning and Sparse Classification Framework, which combines the use of YOLOv8 to extract deep features with L1-regularized machine learning models to do the efficient and interpretable disease recognition.

Contrast limited adaptive histogram equalization (CLAHE) and colour difference histogram (CDH) feature merging capsule network (CCFMCapsNet) for complex image recognition. →

To enhance crop yield, detecting leaf diseases has become a crucial research focus. Deep learning and computer vision excel in digital image processing. Various techniques grounded in deep learning have been utilized for detecting plant leaf diseases; however, achieving high accuracy remains a challenge. Basic convolutional neural networks (CNNs) in deep learning struggle with issues such as the abnormal orientation of images, rotation, and others, resulting in subpar performance. CNNs also need extensive data covering a wide range of variations to deliver strong performance. CapsNet is an innovative deep-learning architecture designed to address the limitations of CNNs. It performs well without needing a vast amount of data in various variations. CapsNets have their limitations, such as the encoder network considering every element in the image and the crowding issue. Due to this, they perform well on simple image recognition tasks but struggle with more complex images. To address these challenges, we introduced a new CapsNet model known as CCFM-CapsNet. This model incorporates CLAHE to reduce image noise and CDH to extract crucial features. Also, max-pooling and dropout layers are incorporated in the original CapsNet model for identifying and classifying diseases in apples, bananas, grapes, corn, mangoes, pepper, potatoes, rice, tomato and also for classifying fashion-MNIST and CIFAR-10 datasets. The proposed CCFM-CapsNet demonstrates significantly high validation accuracies, achieving 99.53%, 95.24%, 99.75%, 97.40%, 99.13%, 100%, 99.77%, 100%, 98.54%, 93.48%, and 82.34% with corresponding parameters in millions(M) 4.68M, 4.68M, 4.68M, 4.68M, 4.79M, 4.63M, 4.66M, 4.68M, 4.84M, 2.39M, and 4.84M for the datasets aforementioned respectively, outperforming the traditional CapsNet and other advanced CapsNet models. Consequently, the CCFM-CapsNet model can be utilized effectively as a smart tool for identifying plant diseases and also in achieving Sustainable Development Goal 2 (Zero Hunger), which aims to end global hunger by the year 2030.

Why it matches plant phenotyping methods植物葉画像から病害状態を推定する深層学習モデルを提案・評価しており、画像取得・分類手法が研究の中心であるため。

abstractTo address these challenges, we introduced a new CapsNet model known as CCFM-CapsNet.

Curvature sensing strategy for flexible gripper fingers during grasping deformation: enabling in-orchard online apple size grading →

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

Why it matches plant phenotyping methods果実を対象に、柔軟グリッパーの変形からリンゴサイズをオンライン推定するセンシング手法が主題であり、植物器官の形態形質測定が中心です。

titleCurvature sensing strategy for flexible gripper fingers during grasping deformation: enabling in-orchard online apple size grading

A Hybrid Approach using ResNet 50 and EfficientNetB0 for Attention-enhanced Deep Learning for Early Detection of Apple Plant Diseases: A Review →

Early diagnosis of plant diseases constitutes a critical determinant in enhancing agricultural productivity and safeguarding global food security, yet current diagnostic methodologies often lack the precision and efficiency required for widespread agricultural implementation. This research presents a novel hybrid deep learning architecture for apple crop disease classification, leveraging the complementary strengths of ResNet50 and EfficientNetB0 frameworks augmented with sophisticated attention mechanisms. The proposed model integrates spatial, channel and custom attention modules to enhance feature extraction capabilities and enable targeted focus on disease-specific regions within plant imagery, representing a significant advancement over our previous MobileNetV2-based implementation which achieved 97% accuracy. The model was trained on an extensive dataset of apple crop images, incorporating advanced data augmentation techniques to improve generalization across diverse environmental conditions and disease manifestations. The hybrid architecture demonstrated superior performance compared to the baseline MobileNetV2 model, achieving a test accuracy of 98.4% with enhanced F1-scores across all disease categories. Comprehensive evaluation through training-validation loss trajectories, receiver operating characteristic curves and confusion matrix analysis confirmed the model’s robustness and clinical efficacy, whilst the attention mechanisms successfully improved the model’s interpretability by highlighting disease-relevant image regions, thereby enhancing diagnostic confidence. The proposed hybrid deep learning model establishes a new benchmark for automated plant disease detection, offering substantial improvements in accuracy and reliability, with future research directions encompassing real-time field deployment and extension to diverse crop species, potentially revolutionizing precision agriculture practices.

Why it matches plant phenotyping methodsリンゴ病害画像から植物の病害状態を推定する深層学習手法を開発し、ベースライン比較と性能評価を行っており、植物フェノタイピング手法が中心である。

abstractThis research presents a novel hybrid deep learning architecture for apple crop disease classification

Apple Plant Disease Classification: Methods, Technologies, and Future Trends →

Apple production, a cornerstone of global agriculture, faces significant threats from diseases such as apple scab, fire blight, powdery mildew, and cedar apple rust, which reduce yield, quality, and sustainability. Early and accurate disease classification is essential to mitigate economic losses and ensure food security. This paper evaluates traditional and modern approaches to apple plant disease classification, including manual visual diagnosis, image-based techniques, and molecular methods like PCR and ELISA. While traditional methods are accessible but error-prone, advanced technologies such as machine learning, deep learning, and sensor-based systems offer high accuracy and scalability, achieving up to 95% detection rates in controlled settings. Challenges, including limited labeled datasets, high computational costs, and poor model generalization across apple varieties and regions, hinder widespread adoption. Emerging trends, such as generative AI, explainable AI, drone-based monitoring, and edge computing, promise to enhance real-time diagnostics and accessibility. The paper also explores opportunities for integrating these technologies with precision agriculture to optimize orchard management and promote sustainability. By synthesizing current methods, technologies, and research gaps, this paper provides a comprehensive roadmap for researchers, farmers, and policymakers to advance apple disease management, fostering sustainable agricultural practices and global food security.

Why it matches plant phenotyping methodsリンゴ植物の病徴・病害状態を画像、センサー、機械学習等で分類する方法を主題としたレビューであり、植物フェノタイピング手法の総説に該当する。

titleApple Plant Disease Classification: Methods, Technologies, and Future Trends

App2: software solution for apple leaf disease detection based on deep learning (CNN+SVM). →

Early detection of crop diseases is essential to reduce yield losses and improve management efficiency in agricultural production. This work presents the development of a mobile application, called App2, designed to detect diseases in apple tree leaves from images taken or uploaded by the user. The solution integrates a hybrid model based on a Convolutional Neural Network (CNN) and a Support Vector Machine (SVM), developed for computer vision tasks focused on recognizing diseases in apple leaves. The system architecture includes a user interface built with React Native, an API developed using FastAPI and deployed on Azure, and a pre-filter implemented through the OpenAI API to validate that the uploaded images correspond to crop leaves. The model was trained to classify images into six categories: Scab, Black Rot, Rust, Healthy, Powdery Mildew, and Spider Mite. Experimental results showed a 95% success rate in test cases and 80% performance in detecting clear images of affected leaves. User evaluations indicated high usability and satisfaction, demonstrating that the mobile application has strong potential as an accessible and effective technological tool for disease monitoring in apple crops.

Why it matches plant phenotyping methodsリンゴ葉の画像から病害状態を推定するCNN+SVMとモバイルアプリを開発しており、植物病害の画像ベース表現型推定が研究の中心である。

abstractThis work presents the development of a mobile application, called App2, designed to detect diseases in apple tree leaves from images taken or uploaded by the user.

BranchMatch: point cloud registration for individual apple trees with limited overlap based on local structure characteristics →

Point cloud registration is a critical technology for 3D reconstruction and personalized management of fruit trees. While ensuring the accuracy and completeness of 3D point cloud reconstruction, the simplest and most efficient approach is to acquire and register point clouds from two stations separated by 180°. For this, we propose BranchMatch, a low-overlap viewpoints acquisition and registration method tailored for tall-spindle individual apple trees during dormancy. The method requires only two point clouds captured from stations 180° apart. Then, it leverages key branch segments in a single viewpoint, utilizing their spatial and geometric structure features in combination with a dynamically weighted feature discriminant function to perform feature matching and initial rigid-body transformation under low overlap conditions. Subsequently, an iterative closest point algorithm, enhanced with local feature matching optimization based on the tree-specific point cloud, is applied to refine the registration and prevent over-registration. Experiments conducted on multiple individual apple trees with two low-overlap point clouds (180° apart) demonstrate a registration success rate of 90%. Compared to the spherical markers registration method, BranchMatch achieves average rotation and translation errors of 1.93 mrad and 4.33 mm, respectively, with a pointwise error of 2.70 mm. Furthermore, compared to multi-site high-overlap registration methods under similar conditions, BranchMatch significantly reduces computational costs while maintaining registration accuracy and reconstruction completeness, highlighting its efficiency and reliability in individual tree registration.

Why it matches plant phenotyping methods個体リンゴ樹の3D点群取得・登録・再構成を目的とする手法を開発し、複数樹体で成功率と誤差を検証しており、植物形態・樹体構造のフェノタイピング基盤として中心的です。

abstractwe propose BranchMatch, a low-overlap viewpoints acquisition and registration method tailored for tall-spindle individual apple trees during dormancy.

A lightweight method for apple disease segmentation using multimodal transformer and sensor fusion →

To address the challenges of multimodal data fusion, low deployment efficiency, and inadequate recognition robustness in complex environments for fruit tree disease segmentation and severity classification, a multimodal parallel transformer-based framework was proposed for apple disease recognition and grading. This method integrates image data with multi-dimensional environmental sensor information. An image segmentation preprocessing module was incorporated to enhance lesion region representation, while a cross-scale attention mechanism and a frame-wise diffusion module were introduced to improve robustness under challenging backgrounds. Additionally, pruning, quantization, and knowledge distillation techniques were employed to enable lightweight deployment. Experimental results demonstrated that the full model achieved outstanding performance on apple disease recognition tasks, reaching a precision of 0.98, recall of 0.93, F1-score of 0.95, and accuracy of 0.96, surpassing several state-of-the-art methods including Mask R-CNN, SegFormer, and Swin Transformer. After compression, the model size was reduced to 76.4 MB, and computational complexity decreased to 6.1 G, enabling real-time inference speeds of 25.2 FPS and 39.6 FPS on Jetson Xavier and Orin platforms, respectively. Ablation studies confirmed the performance contributions of the segmentation preprocessing, sensor fusion, and diffusion modules, demonstrating the potential of the proposed framework for deployment in resource-constrained agricultural scenarios.

Why it matches plant phenotyping methodsリンゴ病斑の画像セグメンテーションと重症度分類を中核とする軽量マルチモーダル手法を開発し、性能比較・アブレーション・実装性能評価まで行っているため、植物フェノタイピング手法として含める。

abstracta multimodal parallel transformer-based framework was proposed for apple disease recognition and grading

Multi leaf disease identification and classification using efficient capsule convolutional shuffle attention network →

Plant diseases establish a serious risk to worldwide food safety and agronomic sustainability, leading to substantial losses in harvest and quality. Precise and appropriate experience of these viruses is necessary for employing effective control procedures, diminishing economic victims and confirming food accessibility. However, current detection methods are often hindered by limitations such as insufficient generalizability across various crops, difficulty in handling complex, and noisy backgrounds, and suboptimal performance when addressing diverse and overlapping disease symptoms. This research is driven by the persistent necessity to tackle these issues and delivers a novel approach to improve disease identification and categorization performance. For this need, propose the Efficient Capsule convolutional Shuffle Attention Network (ECSAN), a comprehensive framework specifically designed to classify diseases across five diverse crops: Apple Leaves, Cassava Leaves, Hibiscus, Hyacinth Bean, and Okra Leaves. The framework integrates a fast gradient-domain weighted guided image filter to denoise and improve image superiority and segmentation is done over a dense swin transformer combined with Unet. Statistical features are extracted using general kernel joint non-negative matrix factorization, and the classification progression is optimized through the Enhanced Osprey Optimization Algorithm (EOOA). Implementation outcomes on five standard datasets prove the superiority of ECSAN, achieving 99.9 % accuracy, 99.98 % recall, 99.97 % precision, 99.98 % F1-score, and 99.99 % specificity. Additionally, ECSAN significantly reduces execution time compared to existing methods, emphasizing its efficiency and applicability in agricultural scenarios. This research underscores the urgency of addressing current detection challenges and establishes a robust foundation for advancements in plant pathology.

Why it matches plant phenotyping methods植物葉の病害状態を画像から抽出・分類する新規深層学習フレームワークを開発し、複数作物・標準データセットで性能評価しているため、植物フェノタイピング手法が中心である。

abstractdelivers a novel approach to improve disease identification and categorization performance

Increasing yield estimation accuracy for individual apple trees via ensemble learning and growth stage stacking →

Accurate prediction of individual apple tree yields during the preharvest stage is essential for precision orchard management and market planning. However, systematic studies focusing on apple yield estimation are scarce. To address this gap, this study targets Fuji apples in the Aksu region of Xinjiang. Multiple images were captured via a UAV during four key growth stages: flowering, fruit formation, fruit expansion, and ripening. On the basis of the extracted vegetation indices, we first developed yield estimation models using random forest (RF), support vector regression (SVR), partial least squares regression (PLS), and ridge regression (RR) methods. We subsequently combined these four models to construct a stacking ensemble learning (SEL) model. To further increase the accuracy of apple yield estimation, we refined the growth stage stacking method and developed a new model, the growth stage stacking ensemble (GSSE). This model maximises the use of spectral information from multiple apple growth stages by employing various machine learning algorithms and integrating multistage spectral data to improve yield estimation accuracy. The results indicate that the optimal period for yield estimation occurs during the fruit expansion stage, with the support vector regression (SVR) model achieving the best performance (R² = 0.654, RMSE = 5.307 kg). Compared with individual machine learning models, the SEL approach enhances yield estimation accuracy, reaching a maximum R² of 0.686 and an RMSE of 5.058 kg. Furthermore, GSSE significantly enhanced accuracy compared with the single-growth stage estimation models and SEL, with the combination of fruit expansion and fruit ripening stages yielding the best results, with an R² of 0.759 and an RMSE of 4.431 kg, with the fruit expansion stage contributing the most. This study is the first to apply the GSSE to apple yield estimation, offering novel insights for UAV-based apple yield estimation.

Why it matches plant phenotyping methodsUAV画像と多時期スペクトル情報から個体別リンゴ収量を推定するモデルを開発・比較しており、植物形質の取得・推定手法が研究の中心である。

abstractMultiple images were captured via a UAV during four key growth stages: flowering, fruit formation, fruit expansion, and ripening.

Classification of Plant Leaf Diseases Using ResNet18 Enhanced with Inception and Capsule Network →

Accurate and early detection of plant leaf diseases is crucial for ensuring crop health and improving agricultural productivity. This work proposes a hybrid deep learning model that combines ResNet18, Inception blocks, and fully connected Capsule layers to classify leaf images of apple, grape, and corn plants into healthy or diseased categories. ResNet18 is used as the backbone for deep feature extraction, while Inception modules enhance the network’s ability to capture multi-scale patterns. Capsule layers are employed at the final stage to retain spatial relationships and pose information, improving the model's ability to recognize complex disease features. The model is trained and evaluated using images from the PlantVillage dataset, with separate configurations for each crop. The proposed model achieved validation accuracies of 99.84% for apple, 100% for grape, and 97.27% for corn. Performance is further assessed using precision, recall, and F1-score, and compared against a baseline ResNet18 model. The results demonstrate that the proposed architecture significantly improves classification accuracy and feature understanding, making it a strong candidate for real-world agricultural disease monitoring systems.

Why it matches plant phenotyping methods植物葉画像から健全・病害状態を推定する画像ベースの深層学習手法を開発・評価しており、病害表現型の抽出が研究の中心である。

abstractThis work proposes a hybrid deep learning model that combines ResNet18, Inception blocks, and fully connected Capsule layers to classify leaf images of apple, grape, and corn plants into healthy or diseased categories.

MemGanomaly: Memory-Augmented Ganomaly for Frost- and Heat-Damaged Crop Detection →

Climate change poses significant challenges to agriculture, leading to increased crop damage owing to extreme weather conditions. Detecting and analyzing such damage is crucial for mitigating its effects on crop yield. This study proposes a novel autoencoder (AE)-based model, termed “Memory Ganomaly,” designed to detect and analyze weather-induced crop damage under conditions of significant class imbalance. The model integrates memory modules into the Ganomaly architecture, thereby enhancing its ability to identify anomalies by focusing on normal (undamaged) states. The proposed model was evaluated using apple and peach datasets, which included both damaged and undamaged images, and was compared with existing robust Convolutional neural network (CNN) models (ResNet-50, EfficientNet-B3, and ResNeXt-50) and AE models (Ganomaly and MemAE). Although these CNN models are not the latest technologies, they are still highly effective for image classification tasks and are deemed suitable for comparative analyses. The results showed that CNN and Transformer baselines achieved very high overall accuracy (94–98%) but completely failed to identify damaged samples, with precision and recall equal to zero under severe class imbalance. Few-shot learning partially alleviated this issue (up to 75.1% recall in the 20-shot setting for the apple dataset) but still lagged behind AE-based approaches in terms of accuracy and precision. In contrast, the proposed Memory Ganomaly delivered a more balanced performance across accuracy, precision, and recall (Apple: 80.32% accuracy, 79.4% precision, 79.1% recall; Peach: 81.06% accuracy, 83.23% precision, 80.3% recall), outperforming AE baselines in precision and recall while maintaining comparable accuracy. This study concludes that the Memory Ganomaly model offers a robust solution for detecting anomalies in agricultural datasets, where data imbalance is prevalent, and suggests its potential for broader applications in agricultural monitoring and beyond. While both Ganomaly and MemAE have shown promise in anomaly detection, they suffer from limitations—Ganomaly often lacks long-term pattern recall, and MemAE may miss contextual cues. Our proposed Memory Ganomaly integrates the strengths of both, leveraging contextual reconstruction with pattern recall to enhance detection of subtle weather-related anomalies under class imbalance.

Why it matches plant phenotyping methodsリンゴ・モモの画像から霜害・高温害という植物の状態を検出する新規異常検出モデルを開発し、複数モデルと比較検証しているため、植物フェノタイピング手法が中心である。

abstractThis study proposes a novel autoencoder (AE)-based model, termed “Memory Ganomaly,” designed to detect and analyze weather-induced crop damage under conditions of significant class imbalance.

Hyperspectral Imaging-Based Deep Learning Method for Detecting Quarantine Diseases in Apples. →

Rapid detection of quarantine diseases in apples is essential for import-export control but remains difficult because routine inspections rely on manual visual checks that limit automation at port scale. A fast, non-destructive system suitable for deployment at customs is therefore needed. In this study, three common apple quarantine pathogens were targeted using hyperspectral images acquired by a close-range hyperspectral camera and analyzed with a convolutional neural network (CNN). Symptoms of these diseases often appear similar in RGB images, making reliable differentiation difficult. Reflectance from 400 to 1000 nm was recorded to provide richer spectral detail for separating subtle disease signatures. To quantify stage-dependent differences, average reflectance curves were extracted for apples infected by each pathogen at early, middle, and late lesion stages. A CNN tailored to hyperspectral inputs, termed HSC-Resnet, was designed with an increased number of convolutional channels to accommodate the broad spectral dimension and with channel and spatial attention integrated to highlight informative bands and regions. HSC-Resnet achieved a precision of 95.51%, indicating strong potential for fast, accurate, and non-destructive detection of apple quarantine diseases in import-export management.

Why it matches plant phenotyping methodsリンゴ病害の症状をハイパースペクトル画像とCNNで検出・識別する方法が研究の中心であり、植物の病害状態を直接推定している。

abstracta convolutional neural network (CNN)

IMNM: integrated multi-network model for identifying pepper leaf diseases. →

As a vegetable crop with high economic value, the yield of pepper is often significantly restricted by leaf diseases, and the spots formed by these diseases on the surface of leaves are highly complex in color and texture characteristics. To overcome the shortcomings of traditional manual identification methods, such as low efficiency, time-consuming, and labor-consuming, an integrated multi-network model (IMNM) was established by combining an improved ResNet, a dynamic convolution network (DCN), and a progressive prototype network (PPN), which was aimed at five typical pepper leaf samples (healthy, virus, leaf blight, brown spot, and phyllosticta). The experimental results show that IMNM achieves 98.55% accuracy in pepper disease identification, which is significantly better than the benchmark models such as Inception-V4, ShuffleNet-V3, and EfficientNet-B7. In the cross-species generalization verification, the average identification accuracy of the model for apple, wheat, and rice leaf diseases increased to 99.81%, and its four core indicators of specificity, precision, sensitivity, and accuracy were all stable over 98%. This demonstrates that IMNM can effectively analyze the color and texture characteristics of highly heterogeneous disease spots and possesses strong cross-crop generalization capabilities. Its technical path lays a theoretical foundation for the development of field mobile disease diagnosis equipment based on deep learning, and is of great value for promoting the engineering application of an intelligent monitoring system for crop diseases and insect pests.

Why it matches plant phenotyping methods葉表の病斑を画像から識別する深層学習モデルを開発・比較評価しており、植物病害状態の取得・推定手法が中心である。

abstractan integrated multi-network model (IMNM) was established by combining an improved ResNet, a dynamic convolution network (DCN), and a progressive prototype network (PPN)

Apple leaf disease image recognition based on a modified rime optimization algorithm and ConvNeXt network. →

Early and accurate diagnosis of apple leaf disease is a prerequisite for maintaining crop health and for enhancing agricultural productivity. Conventional methods, which largely relied on human inspection or naive machine learning algorithms, were not capable of handling the complexity of patterns, the class imbalance, and the real-world challenges such as conflated symptoms or poor lighting. The present study develops a completely new model design by integrating a ConvNeXt model along with a modified rime optimization algorithm (MRIME) used for hyperparameter tuning as well as complementing through the Convolutional Block Attention Module (CBAM) to ensure better feature extraction. CBAM extends the power of the model in focusing on critical discriminative regions, while MRIME gives optimal values for relevant hyperparameters for generalization while avoiding overfitting. Evaluated by the Apple Leaf Disease Symptoms Dataset, the proposed approach attained an accuracy of 92.7%, precision of 92.5%, recall of 92.6%, F1-score of 92.5%, and mAP of 92.3%, surpassing most baselines including ResNet50 and EfficientNet-B0. Compared to the aforementioned baselines, ablation experiments demonstrated that CBAM led to about 1.5% enhancement in accuracy, while MRIME could boost performance by another 1.2% via hyperparameter tuning. These results confirm the complementary benefit of attention mechanisms and metaheuristic optimization in producing state-of-the-art results.

Why it matches plant phenotyping methodsリンゴ葉の病徴を画像から分類・認識する計算手法を開発し、データセット上で性能評価・比較しており、植物の病害状態推定が方法論の中心である。

abstractThe present study develops a completely new model design by integrating a ConvNeXt model along with a modified rime optimization algorithm (MRIME) used for hyperparameter tuning as well as complementing through the Convolutional Block Attention Module (CBAM) to ensure better feature extraction.

Online multi-view multispectral detection for early bruised apple →

Online multispectral dynamic inspection is crucial for smart agriculture, particularly in acquiring multispectral image data across the entire surface of fruits during the inspection process. This study focuses on early bruises in apples, presenting an online multispectral multi-surface imaging strategy. The proposed strategy is based on an imaging model using two side mirrors, combined with an imaging sensor with a lens-filter array. This configuration enables the rapid capture of spatial texture and multispectral information from the multiple viewing directions for a sample in a single imaging process of one CCD. During the design process, a monochromatic LED-based integrating sphere optical system is introduced to uniformly illuminate the entire surface of the apple samples. Based on this, a mathematical model is established for the side mirror layout and system geometric parameters to determine the system configuration that scans the sample surface. In practical applications, the proposed method achieved an effective classification rate of 91 % for three quality categories of apples—sound, slightly bruised, and severely bruised—at a detection speed of about 3 per second. These results suggest that this study provides potential technical support for apple quality monitoring in smart agriculture.

Why it matches plant phenotyping methodsリンゴ果実の打撲状態を対象に、多視点マルチスペクトル撮像システムと画像取得戦略を設計・評価しており、植物状態の取得方法が研究の中心である。

abstractpresenting an online multispectral multi-surface imaging strategy

RTFVE-YOLOv9: Real-time fruit volume estimation model integrating YOLOv9 and binocular stereo vision →

This study proposes a real-time fruit volume estimation model based on YOLOv9 (RTFVE-YOLOv9) and binocular stereo vision technology to address the challenges of low automation and insufficient accuracy in fruit volume measurement in complex orchard environments, particularly in scenarios with diverse canopy structures and severe branch-leaf occlusion. The model achieves effective recognition of occluded fruits through the innovative design of a Dual-Scale and Global–Local Sequence (DSGLSeq) module while incorporating a Multi-Head and Multi-Scale Self-Interaction (MHMSI) module to improve the detection performance of small fruit targets. Systematic validation experiments conducted on major economic fruit tree varieties, including apples, pears, pomelos, and kiwifruit, demonstrate that RTFVE-YOLOv9 improved the mean Average Precision (mAP) by 2.1%, 1.6%, 4%, and 3.8% respectively on the four fruit datasets compared to the baseline YOLOv9-c model. The model’s internal working mechanisms were thoroughly revealed through multi-dimensional evaluation, including ablation experiments, Heatmap Analysis, and Effective Receptive Field (ERF) analysis, providing a theoretical foundation for subsequent optimization. The research findings enrich the application theory of computer vision in smart agriculture and provide reliable technical support for achieving precise orchard management.

Why it matches plant phenotyping methods果実の体積という植物器官形質を、YOLOv9と両眼ステレオビジョンで推定する手法を開発・検証しており、画像取得・計算による表現型推定が研究の中心である。

abstractThis study proposes a real-time fruit volume estimation model based on YOLOv9 (RTFVE-YOLOv9) and binocular stereo vision technology

Foundation model-based apple ripeness and size estimation for selective harvesting →

Harvesting is a critical task in the tree fruit industry, demanding extensive manual labor and substantial costs, and exposing workers to potential hazards. Recent advances in automated harvesting offer a promising solution by enabling efficient, cost-effective, and ergonomic fruit picking within tight harvesting windows. However, existing harvesting technologies often indiscriminately harvest all visible and accessible fruits, including those that are unripe or undersized. This study introduces a novel foundation-model-based framework for efficient apple ripeness and size estimation. Specifically, we curated two public RGBD-based Fuji apple image datasets, integrating expanded annotations for ripeness (“Ripe” vs. “Unripe”) based on fruit color and image capture dates. The resulting comprehensive dataset, Fuji-Ripeness-Size Dataset, includes 4,027 images and 16,257 annotated apples with ripeness and size labels. To the best of our knowledge, this is the first published dataset on apples with ripeness and size annotations. Leveraging Grounding-DINO, a foundation-model-based object detector, we achieved robust apple detection and ripeness estimation, with mean Average Precision being 72.8, outperforming other state-of-the-art models in the evaluation on our dataset. Additionally, we developed six size estimation algorithms, made a comprehensive comparison using box-plots, and identified the best algorithm with lowest error and variation. The Fuji-Ripeness-Size Dataset and the apple detection and size estimation algorithms are made publicly available¹1The code and dataset is available at https://github.com/zhukeyi-stan/Fuji_Ripeness_And_Size_Estimation., which provides valuable benchmarks for future studies in automated and selective harvesting.

Why it matches plant phenotyping methodsリンゴの熟度・サイズという植物器官の形質を画像から推定する手法を開発・比較し、データセットとベンチマークも提供しているため、フェノタイピング手法が中心である。

abstractThis study introduces a novel foundation-model-based framework for efficient apple ripeness and size estimation.

Plant leaf disease detection and classification using CNN and VGG16 models →

Our lives are greatly impacted by the agricultural sector.The most significant industry in our economy is agriculture.The result of effective management is a successful agricultural product.Farmers that are unaware of leaf disease produce less.Profit and loss are determined by production, hence identifying plant leaf diseases is essential.The solution for categorizing and identifying leaf diseases is CNN.This study aims to identify leaf diseases in potato, tomato, corn, grape, and apple plants.Large agricultural disease monitoring fields are monitored for plant leaf diseases, which automatically identify certain disease characteristics and cure them.Comparing the proposed CNN model to popular transfer learning methods such as VGG16.There are numerous applications for plant leaf disease detection across a range of sectors, including biological research and agricultural institutions.One of the necessary study topics is plant leaf disease detection since it may help monitor vast agricultural fields and automatically identify disease symptoms.

Why it matches plant phenotyping methods植物葉の病徴を画像からCNN/VGG16で検出・分類する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 study aims to identify leaf diseases in potato, tomato, corn, grape, and apple plants.

Development of a Machine vision system for apple bud thinning in precision crop load management →

Thinning is a critical practice in apple orchard management, directly influencing crop load and fruit quality. To assist automated crop load management, a machine vision system for apple bud detection was developed to be integrated with robotic platforms. The system employed a Kinect Azure sensor for real-time bud detection and branch diameter measurement, utilizing a YOLOv8-based object detection model trained and evaluated across multiple datasets. The evaluation identified the best-performing model by balancing precision, recall, and robustness in the complex and unstructured environments of apple orchards. Several training configurations were assessed, with the selected setup demonstrating a strong balance between precision (68 %), recall (55 %), F1-score (61 %), and mean average precision (mAP: 59 %) across diverse and unstructured orchard environments. This configuration, trained on a combination of FLIR and Kinect Azure data, was chosen for deployment due to its robustness and compatibility with the Kinect Azure sensor in real-world applications. Two proposed imaging methods for branch diameter measurement were validated against manual caliper-based measurements, with statistical analysis revealing no significant differences (p = 0.98). These findings confirm the semi-automated methods as reliable and labor-efficient alternatives for field applications. Additionally, the bud counting algorithm demonstrated accurate tracking and counting of apple buds, effectively avoiding omissions and duplications in real orchard settings. This study underscores the potential of vision systems to revolutionize apple bud thinning, providing a strong foundation for the development of fully automated solutions in precision orchard management.

Why it matches plant phenotyping methodsリンゴ芽の画像検出に加え、枝径という植物形質の画像計測法を開発・手動測定と検証しており、フェノタイピング手法が中心である。

abstracta machine vision system for apple bud detection was developed to be integrated with robotic platforms

DATR: Diffusion-based 3D Apple Tree Reconstruction Framework with Sparse-View →

Digital twin applications offered transformative potential by enabling real-time monitoring and robotic simulation through accurate virtual replicas of physical assets. The key to these systems is 3D reconstruction with high geometrical fidelity. However, existing methods struggled under field conditions, especially with sparse and occluded views. This study developed a two-stage framework (DATR) for the reconstruction of apple trees from sparse views. The first stage leverages onboard sensors and foundation models to semi-automatically generate tree masks from complex field images. Tree masks are used to filter out background information in multi-modal data for the single-image-to-3D reconstruction at the second stage. This stage consists of a diffusion model and a large reconstruction model for respective multi view and implicit neural field generation. The training of the diffusion model and LRM was achieved by using realistic synthetic apple trees generated by a Real2Sim data generator. The framework was evaluated on both field and synthetic datasets. The field dataset includes six apple trees with field-measured ground truth, while the synthetic dataset featured structurally diverse trees. Evaluation results showed that our DATR framework outperformed existing 3D reconstruction methods across both datasets and achieved domain-trait estimation comparable to industrial-grade stationary laser scanners while improving the throughput by $\sim$360 times, demonstrating strong potential for scalable agricultural digital twin systems.

Why it matches plant phenotyping methodsリンゴ樹の疎視点画像から3D形状を再構成し、樹体形質を推定する手法の開発と、圃場・合成データでの評価が研究の中心であるため。

abstractThis study developed a two-stage framework (DATR) for the reconstruction of apple trees from sparse views.

Comparative Analysis of Conventional and Deep Learning Algorithms for Apple Detection →

Abstract Precise apple counting is one of the vital tasks in the overall context of agricultural applications, from yield estimation to resource allocation and several other logistical issues about harvesting. The paper has discussed a comparison of conventional image processing techniques with modern deep learning for the detection of apples. Apple detection research, focusing on apples under difficult orchard situations using the MinneApple dataset to ensure strong deep learning-based detection due to YOLOv8, was studied. In particular, it compared traditional approaches using HSV color segmentation and morphological operations against a fine-tuned YOLOv8 model, improved by Roboflow. These studies proved the deep learning approach of difficult scenarios with very high precision, recall, and F1 score, thus enabling it to rightly distinguish apples on the trees from those lying on the ground. Results are highly useful in understanding how traditional and modern methods can be integrated into agricultural automation.

Why it matches plant phenotyping methodsリンゴ果実の画像検出・計数手法を従来法とYOLOv8で比較評価しており、果実数という植物器官形質の抽出が中心的な技術貢献である。

abstractThe paper has discussed a comparison of conventional image processing techniques with modern deep learning for the detection of apples.

Multi-kernel inception-enhanced vision transformer for plant leaf disease recognition →

Abstract The timely and precise identification of diseases in plants is essential for efficient disease control and safeguarding of crops. Manual identification of diseases requires expert knowledge in the field, and finding people with domain knowledge is challenging. To overcome the challenge, computer vision-based machine learning techniques have been proposed by the researchers in recent years. Most of these solutions with the standard convolutional neural network (CNN) approaches use uniform background laboratory setup leaf images to identify the diseases. However, only a few works considered real-field images in their work. Therefore, there is a need for a robust CNN architecture that can identify the diseases in plants in both laboratory and real-field conditioned images. In this paper, we have proposed an Inception-Enhanced Vision Transformer (IEViT) architecture to identify diseases in plants. The proposed IEViT architecture extracts local as well as global features, which improves feature learning. The use of multiple filters with different kernel sizes efficiently uses computing resources to extract relevant features without the need for deeper networks. The robustness of the proposed architecture is established by hyper-parameter tuning and comparison with state-of-the-art. In the experiment, we consider five datasets with both laboratory-conditioned and real-field conditioned images. From the experimental results, we see that the proposed model outperforms state-of-the-art deep learning models with fewer parameters. The proposed model achieves an accuracy rate of 99.23% for the apple leaf dataset, 99.70% for the rice dataset, 97.02% for the ibean dataset, 76.51% for the cassava leaf dataset, and 99.41% for the plantvillage dataset.

Why it matches plant phenotyping methods植物葉の病害状態を画像から認識する新規深層学習手法を提案・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。

abstractIn this paper, we have proposed an Inception-Enhanced Vision Transformer (IEViT) architecture to identify diseases in plants.

PLANXMAMBA: A hybrid CNN – MAMBA model for plant disease recognition →

Abstract Image-based plant disease recognition plays a pivotal role in smart agriculture , facilitating early detection and effective management of crop diseases. Existing approaches primarily employ Convolutional Neural Networks (CNNs) or Vision Transformers (ViTs) to extract discriminative visual features and perform disease classification, achieving encouraging outcomes. However, these models often struggle to adequately model long-range spatial dependencies and sequence-level information while maintaining lightweight architectures suitable for deployment on mobile or edge devices. In this study, we introduce Plan-tXMamba, an efficient hybrid model that synergistically integrates CNNs with a structured State Space Model (SSM), termed Mamba, to simultaneously capture local and global contextual features. The architecture is designed to enhance accuracy, interpretability, and computational efficiency. Comprehensive experiments were conducted on multiple benchmark datasets—including Maize, Rice, Apple, Embrapa, and PlantVillage—to demonstrate the effectiveness and generalizability of the proposed approach.

Why it matches plant phenotyping methods植物病害状態を画像から認識する新規CNN–Mamba手法を開発し、複数ベンチマークで有効性・汎化性を検証しており、植物フェノタイピング手法が中心である。

abstractImage-based plant disease recognition plays a pivotal role in smart agriculture

Apple leaf disease severity grading based on deep learning and the DRL-Watershed algorithm. →

Apple leaf diseases significantly impair the photosynthetic efficiency and growth quality of apple trees, leading to reduced fruit yields. Existing methods for disease detection and severity classification struggle to quickly and accurately segment and quantify diseased areas on leaves, particularly in complex backgrounds. To address this issue, we propose a method for assessing the severity of apple leaf diseases based on a combination of improved HRNet and DRL-watershed algorithms. First, we selected HRNet_w32 as the backbone feature extraction network and incorporated a Normalization Attention Mechanism (NAM). Then, we combined the Dice Loss and Focal Loss functions to construct an enhanced HRNet based semantic segmentation model for pixel-level segmentation of both apple leaf and diseased regions. Furthermore, the segmented leaf and disease regions were further optimized using the DRL-watershed algorithm to distinguish overlapping leaf regions. Experimental results demonstrate that the modified HRNet model achieved a mean intersection over union (mIoU) of 88.91% and a mean pixel accuracy (mPA) of 94.13%, representing improvements of 8.77 and 7.25% points, respectively, over the original HRNet. The disease severity assessment accuracy reached 97.65%. This study not only accurately segments apple leaves and diseased areas, but also effectively addresses the impact of complex backgrounds and leaf overlap on disease severity assessment, providing a solid scientific basis for disease management strategies.

Why it matches plant phenotyping methodsリンゴ葉の病斑領域を画像から分割・定量し、病害重症度を推定する深層学習・DRL-Watershed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。

abstractExisting methods for disease detection and severity classification struggle to quickly and accurately segment and quantify diseased areas on leaves

An improved DSCCA-UNet for apple leaf disease severity estimation and prescription map generation →

Abstract This paper investigates the problems of disease severity estimation and prescription map generation. By introducing dynamic snake convolution (DSC), a DSCC multi-scale feature extraction module is designed to build a new UNet architecture. The combination of DSCC module with VGG16 backbone can enhance the receptive field for segmented edge feature information which can achieve more detailed edge feature fusion. The channel attention (CA) and spatial attention (SA) in the CBAM attention module are disassembled and used in the skip connection part and the upsampling part, respectively. This new connection can obtain more location information of the spots and mitigate the effect of background on network learning. Moreover, an automatic pixel counting algorithm based on the improved DSCCA-UNet is designed to estimate the disease severity. Finally, the system of the apple leaf disease severity estimation and the variable prescription maps are obtained based on the PyQt5 tool and ArcGIS component. The experimental results show that the improved DSCCA-UNet model outper-forms other mainstream semantic segmentation models. It can more effectively complete the tasks of disease severity estimation and prescription map generation in actual orchard scenarios.

Why it matches plant phenotyping methodsリンゴ葉の病斑を画像分割・画素計数し、植物体の病害重症度を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。

abstractan automatic pixel counting algorithm based on the improved DSCCA-UNet is designed to estimate the disease severity.

Dataset of apples for grading by sweetness, ripeness and variety →

The study created a detailed database for apple quality inspection using a cost-effective, self-designed multi-spectral imaging system. The system was optimized to allow spectral information to be obtained in 8 discrete wavebands, which enabled non-destructive determination of such key fruit components as ripeness, sugar type and cultivar. Stringent environmental conditions were maintained during image acquisition for optimal measurement consistency and experimental repeatability. The detailed dataset encompasses 32,463 multi-spectral images across three distinct classification categories. For sweetness evaluation, 1620 images spanning Brix values from 10 % to 15 % were collected from five apple varieties. Ripeness evaluation includes 29,160 images documenting the complete maturation cycle over 18 days, while variety classification contains 1683 images from three distinct cultivars. Each image was captured under controlled lighting conditions using eight specific wavelengths, ensuring spectral consistency crucial for machine learning applications. These multi-spectral images were concatenated for grading by sweetness, ripeness, and variety, creating a processed dataset of concatenated images optimized for AppleNet processing. The concatenation process combines the eight wavelength channels into unified image representations suitable for deep learning applications. Sample collection included the picking of different apple cultivars at different physiological development phases of fruit from local orchards. Single specimens were imaged sequentially using a multi-spectral technique. Information on sugar content concentration (% Brix), maturation phase classification and varietal identification was recorded according to standard laboratory procedure. The resulting annotated database includes such quantitative reference points, which can be used to train supervised learning classifiers in computational classification systems. The reuse value of the dataset covers a wide range of applications such as machine learning-based fruit quality evaluation, agricultural automation and food industry examination. This dataset of ours can be used by researchers to develop and test algorithms to classify apples and estimate their ripeness and the presence of diseases. Furthermore, the proposed multi-spectral imaging can be generalized to cover other fruits and agricultural products, extending the application of the method in smart agriculture. This dataset serves as a valuable resource for researchers in computer vision, machine learning, and agricultural technology, fostering advancements in non-destructive fruit quality evaluation methodologies.

Why it matches plant phenotyping methodsリンゴ果実の糖度・成熟度・品種という植物器官の形質を対象に、マルチスペクトル撮像システムを構築・最適化し、注釈付き大規模データセットを作成しているため、フェノタイピング手法とデータセットが中心です。

abstractThe study created a detailed database for apple quality inspection using a cost-effective, self-designed multi-spectral imaging system.

Explainable AI Meets MobileNetV2: A Multi-Branched Approach for Apple Leaf Disease Identification →

Apple farming plays a significant role in agriculture, serving as an essential source of livelihood for farmers. However, cedar apple rust, apple scab, and black rot are common apple leaf diseases severely affecting apple yield. Early detection of these diseases is crucial for preserving quality and productivity. Researchers have used deep learning models to improve disease classification, they but lack lightweight architecture and transparency, operating as black box systems. Leveraging the power of lightweight models and explainable artificial intelligence (XAI) addresses these challenges by developing transparent methods based on convolutional neural networks (CNNs). This study proposes an enhanced version of MobileNetV2, incorporating a multi-branched architecture to improve feature map representation for classification tasks. The proposed model achieved 99.18% accuracy on the benchmark plant village dataset, surpassing existing studies. The results also integrated local interpretable model-agnostic explanations (LIME) to emphasize the role of individual features in the model’s predictions. The proposed model has a lightweight structure, which ensures its suitability for IoT-based real-time agricultural applications.

Why it matches plant phenotyping methodsリンゴ葉の病害状態を画像から分類する軽量CNNと説明可能AI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 study proposes an enhanced version of MobileNetV2, incorporating a multi-branched architecture to improve feature map representation for classification tasks.

Multi-architecture CNN Framework with MBConv-SE Optimization for Fruit Disease Classification →

Accurate fruit disease classification is critical for agricultural productivity, yet traditional manual methods remain labour-intensive and error-prone. This study proposes a hybrid model that synergistically integrates multi-architecture convolutional neural network models such as ResNet, DenseNet, and EfficientNet to address challenges such as intra-class variability and dataset imbalances. The model has Mobile Inverted Bottleneck Convolution with Squeeze-and-Excitation blocks that help it prioritize useful features. It also has residual pathways for gradient optimization and data augmentation that makes it easier to generalize. The results are evaluated on 9714 apple disease images (four classes), 3403 guava disease images (three classes), and 5000 mango disease images (five classes), the hybrid model achieves perfect performance, with 99.99% accuracy, precision, recall, and F1-score for apple diseases, 99.60% accuracy for guava diseases, and 99.68% accuracy for mango diseases, surpassing state-of-the-art methods. Key innovations include dense connectivity for feature reuse, channel-wise attention mechanisms, and computational efficiency through global average pooling. The results underscore the model’s robustness and scalability, offering a practical solution for real-time disease detection in smart farming systems.

Why it matches plant phenotyping methods果実の病害状態を画像から分類するCNN手法の開発と評価が中心であり、植物の病徴・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収録する。

abstractThis study proposes a hybrid model that synergistically integrates multi-architecture convolutional neural network models such as ResNet, DenseNet, and EfficientNet

ALD-YOLO: a lightweight attention detection model for apple leaf diseases. →

As an important economic crop, apples are significantly affected by disease infestations, which can lead to substantial reductions in apple yield and economic losses. To rapidly and accurately detect apple leaf diseases, we propose a lightweight attention detection model ALD-YOLO based on the YOLOv8 architecture. To improve overall efficiency, we design the Faster_C2F module within the Backbone and Neck by optimizing YOLOv8's primary C2F (Faster Implementation of CSP Bottleneck with 2 convolutions) modules with the more computationally effective FasterNet Block. To strengthen the model's ability to capture multi-scale feature information and focus on smaller disease targets, the EMA (Efficient Multi-Scale Attention) module is introduced at the input end where the Neck connects to the detection module of the Head, forming a new Faster_C2F_EMA module. Two novel C2F modules can achieve the optimal balance of detection accuracy and efficiency. Furthermore, to reduce the model's parameters and retain more image information, most convolution modules in the YOLOv8 architecture are replaced by a lightweight downsampling module ADown. In comparison with YOLOv8n and YOLOv8s, experimental results on the AppleLeaf9 dataset showed that ALD-YOLO increased mAP by 1.4% and 0.6%, and reduced GFLOPs by 29.63% and 79.93%, respectively. The CPU inference testing showed that the improvement of our model in frames per second reached up to 119.23% compared to YOLOv8s. Therefore, our model delivers more stable and efficient detection of apple leaf diseases, even on edge devices.

Why it matches plant phenotyping methodsリンゴ葉の病徴を画像から検出・分類する軽量モデルを開発し、精度・計算量・推論速度を比較評価しているため、植物病害状態の画像ベース表現型抽出が中心です。

abstractTo rapidly and accurately detect apple leaf diseases, we propose a lightweight attention detection model ALD-YOLO based on the YOLOv8 architecture.

Quantifying the severity of Marssonina blotch on apple leaves: development and validation of a novel spectral index. →

Apple Marssonina blotch (AMB) is a major disease causing pre-mature defoliation. The occurrence of AMB will lead to serious production decline and economic losses. The precise identification of AMB outbreaks and the measurement of its severity are essential for limiting the spread of the disease, yet this issue remains unaddressed to this day. Given these, we conducted experiments in Qian County, Shaanxi, China, to develop an Apple Marssonina Blotch Index (AMBI) based on hyperspectral imaging, aimed to quantify disease severity at the leaf scale and to monitor infection at the canopy scale. Based on the separability and combination of individual band, characteristic wavelengths were identified in green band, red edge band and near-infrared band to construct AMBI = (R 762nm - R 534nm )/(R 534nm + R 690nm ). The results demonstrated that AMBI exhibited high overall accuracies (R 2 = 0.89, RMSE = 9.67%) in estimating the disease ratio at the leaf scale compared to commonly used indices. At the canopy scale, AMBI enabled effective classification of healthy and diseased trees, yielding an overall accuracy (OA) of 89.09% and a Kappa coefficient of 0.78. Furthermore, analysis of unmanned aerial vehicle (UAV) acquired hyperspectral imagery using AMBI enabled the spatial mapping of diseased tree distribution, highlighting its potential as a scalable and timely tool for precision orchard disease surveillance.

Why it matches plant phenotyping methodsリンゴ葉の病害重症度という植物状態を、ハイパースペクトル画像から定量推定する指標を開発・検証しており、フェノタイピング手法が中心である。

abstractwe conducted experiments in Qian County, Shaanxi, China, to develop an Apple Marssonina Blotch Index (AMBI) based on hyperspectral imaging, aimed to quantify disease severity at the leaf scale and to monitor infection at the canopy scale.

Comprehensive Study of Plant Leaf Disease Detection and Classification using Deep Learning Approach →

Plant leaf diseases pose a major challenge in the agricultural field, leading to poor crop performance and impacting food safety. Timely identification and accurate classification of diseases can prevent plant damage and ensure sustainable farming practices. This paper demonstrates deep learning (DL) approaches for plant leaf disease detection and classification, focusing on Convolutional Neural Networks (CNN) and hybrid/ensemble learning methods. Various datasets, including PlantVillage, Rice Leaf Disease, Corn Leaf Disease, and Apple Leaf Disease, are used for training and testing the models. CNN-based methods such as VGG16, MobileNet, and CapsNet are employed for automatic feature extraction, providing high accuracy. Hybrid models, combining CNN with other algorithms like RNN or SVM, aim to improve performance by addressing issues such as overfitting. A comparative analysis of existing methods is provided, highlighting their advantages, limitations, and performance metrics. The performance of the models is evaluated based on metrics such as accuracy, precision, recall, and F1-score. This paper aims to assist researchers in understanding computer vision applications for plant leaf disease classification.

Why it matches plant phenotyping methods植物葉の病害状態を画像から検出・分類する深層学習手法を比較評価しており、病害表現型の取得・推定が中心です。

abstractThis paper demonstrates deep learning (DL) approaches for plant leaf disease detection and classification, focusing on Convolutional Neural Networks (CNN) and hybrid/ensemble learning methods.

OrchardQuant ‐ 3D : combining drone and LiDAR to perform scalable 3D phenotyping for characterising key canopy and floral traits in fruit orchards →

Summary Orchard fruits such as pear and apple are important for ensuring global food security and agricultural economy as they not only provide essential nutrients, but also support biodiversity and ecosystem services. Breeders, growers and plant researchers constantly study desirable tree morphological features and floral characteristics to ensure fruit production and quality. Still, traditional orchard phenotyping is often laborious, limited in scale and prone‐to‐error, resulting in many attempts to develop reliable and scalable toolkits to address this challenge. Here, we present OrchardQuant‐3D, an analytic pipeline for automating tree‐level analysis of key canopy and floral traits for different types of fruit orchards. We first built a data fusion algorithm to register 3D point clouds collected by both drones (for colour signals) and Light Detection And Ranging (LiDAR, for precise spatial properties), reconstructing high‐quality 3D orchard models at different growth stages. Then, we utilised precise global navigation satellite system signals to position trees in orchards with millimetre‐level accuracy, enabling tree‐level analysis of key canopy (e.g. crown volume and the number or branches) and floral traits (e.g. blossom clusters and volumes) using 3D computer vision, complex graph theory and feature engineering techniques. Equipped with the OrchardQuant‐3D pipeline, we successfully measured varietal differences of four pear cultivars from a small pear orchard in Nanjing China, followed by a scale‐up study that surveyed 3D tree morphologies, key floral and fruit traits from 1104 apple trees in an orchard in East Malling, United Kingdom. To the best of our knowledge, such a multi‐source, comprehensive and expandable methodology has not yet been introduced to this important research domain. Hence, we believe that our work demonstrates a step change in our ability to conduct scalable 3D orchard phenotyping, which is highly valuable to advance orchard breeding, precise tree management and orchard research greatly to sustain fruit tree production in a rapidly changing climate.

Why it matches plant phenotyping methodsドローンとLiDARのデータ融合、3D再構成、コンピュータビジョンによる樹冠・花形質の自動抽出パイプラインを開発しており、植物フェノタイピング手法が研究の中心である。

abstractHere, we present OrchardQuant‐3D, an analytic pipeline for automating tree‐level analysis of key canopy and floral traits for different types of fruit orchards.

Real-time Identification and Quantification of Apple Scab on Fruit in Preharvest and Postharvest Conditions Using YOLOv11: A Deep Learning Approach →

Abstract Background Apple scab (AS), caused by the fungal pathogen Venturia inaequalis , is a major disease of apple that manifests as lesions on leaves and fruits. It significantly reduces fruit quality and yield, leading to substantial economic losses. Traditional AS assessment relies on visual scoring, which is labor-intensive, subjective, and poorly reproducible. This study proposes a deep learning-based framework to overcome these limitations and to enable an accurate, scalable AS phenotyping approach. Results Deep learning techniques were employed for the object detection and segmentation of AS symptoms in apple fruits. A two-stage fine-tuning process using the YOLO foundation model (YOLOv11) was applied to color images collected under orchard and laboratory conditions. The first model achieved over 90% precision in detecting apples, while the second achieved 78% precision in identifying and quantifying AS lesions. The YOLO-based architecture supports the real-time processing of both images and video streams, enabling rapid in situ evaluation. Despite challenges such as variable lighting, shading, and symptom heterogeneity across developmental stages, the model’s performance was enhanced through extensive data augmentation, a diverse image dataset, and the use of high-resolution (840 × 840 pixels) training images, which improved detection of fine-scale features by 40%. Compared to manual scoring, this method is significantly faster, more objective, and more reproducible. Conclusion These results demonstrate the strong potential of the proposed deep learning-based approach as a robust and scalable tool for automated AS phenotyping. By improving the precision and efficiency of disease assessments in both controlled and field environments, this framework effectively supports apple grading assessments and accelerates breeding efforts aimed at identifying AS-resistant genotypes. Moreover, it establishes a solid foundation for broader applications in real-time plant disease monitoring and the future integration of additional apple diseases.

Why it matches plant phenotyping methodsリンゴ果実上の黒星病病斑を画像から検出・セグメンテーションし、病斑を定量化する深層学習法の開発であり、植物病害表現型の取得が研究の中心である。

abstractThis study proposes a deep learning-based framework to overcome these limitations and to enable an accurate, scalable AS phenotyping approach.

Nitrogen content estimation of apple trees based on simulated satellite remote sensing data. →

Introduction Using satellite remote sensing technology to diagnose apple tree nitrogen content is critical for guiding regional precision fertilization of apple trees. However, due to differences in spatial resolution and spectral response, there is a lack of systematic evaluation of satellite data's applicability and accuracy in apple tree nitrogen inversion. Methods This study used apple orchards in Qixia City, Shandong Province as the research area, collecting canopy hyperspectral data through an ASD spectrometer during three key phenological periods: the new-shoot-growing stage (NGS), the new-shoot-stop-growing stage (NSS), and the autumn shoot-growing stage (ASS). The data was resampled based on satellite sensor spectral response functions to match the band resolutions of multiple satellite sources. Correlation coefficient method and partial least squares regression were used to screen sensitive bands for apple tree nitrogen content. Support Vector Machine (SVM) and Backpropagation Neural Network (BPNN) algorithms were used to construct and screen the optimal models for apple tree nitrogen content estimation. Results Results showed that visible light, red edge, near-infrared, and yellow edge bands were sensitive bands for estimating apple tree nitrogen content. The support vector machine model constructed based on Sentinel-2 satellite simulated data was the optimal nitrogen content inversion model, with an average R ² value of 0.81 and an average RMSE value of 0.15 for training sets across different phenological periods, and an average R² value of 0.61 and an average RMSE value of 0.23 for validation sets. Discussion This study systematically evaluated the applicability and accuracy differences of multi-source satellite data for estimating nitrogen content in apple trees, and clarified the variation patterns of nitrogen-sensitive spectral bands and optimal modeling strategies across key phenological stages. This research provides a scientific basis for data selection and a technical paradigm for remote sensing-based nutrient diagnosis of apple trees at the regional scale, and holds significant theoretical and practical value for developing region-wide precision fertilization systems based on remote sensing.

Why it matches plant phenotyping methodsリンゴ樹の窒素含量という植物形質を、模擬衛星スペクトルと機械学習で推定する手法を開発・比較・検証しており、フェノタイピング手法が中心である。

abstractsystematically evaluated the applicability and accuracy differences of multi-source satellite data for estimating nitrogen content in apple trees

Panoptic segmentation for complete labeling of fruit microstructure in 3D micro-CT images with deep learning. →

Metabolic processes in plant organs involving transport of water, metabolic gasses, and nutrients depend on the three-dimensional (3D) microscopic tissue morphology. However, imaging and quantifying this microstructure, including the spatial layout of parenchyma cells, pores, vascular bundles and special features such as stone cell clusters (brachysclereids), is challenging. To address this, a 3D deep learning-based panoptic segmentation model, combining semantic and instance segmentation, was developed to accelerate and improve microstructure characterization of apple and pear fruit tissue in X-ray micro-computed tomography (CT) images. In addition, various training datasets and data augmentation techniques, including synthetic data, were explored to enhance segmentation quality. The 3D panoptic segmentation achieved an Aggregated Jaccard Index of 0.89 and 0.77 for apple and pear tissue, respectively, outperforming both the previously designed 2D instance segmentation model and a marker-based watershed segmentation benchmark. The model successfully labeled vascular bundles with a Dice Similarity Coefficient (DSC) of 0.51 in apple tissue and 0.79 in pear tissue, although thin vasculature in apple remained more challenging to segment. The 3D panoptic segmentation model achieved a DSC of 0.81 and effectively segmented stone cell clusters in pear tissue. Despite evaluating different methods to enhance segmentation quality, none improved test performance beyond that of the model trained on the standard dataset. The proposed 3D panoptic segmentation model offers the most complete automated protocol to date for plant tissue labelling and morphometric quantification from native X-ray micro-CT images, without extensive sample preparation such as contrast labelling. The developed method, if not replaces, drastically accelerates conventional human-in-the-loop analysis of such images.

Why it matches plant phenotyping methods植物組織の3DマイクロCT画像から微細構造を自動セグメンテーションし、形態計測する手法の開発・比較検証が中心であるため。

abstracta 3D deep learning-based panoptic segmentation model, combining semantic and instance segmentation, was developed to accelerate and improve microstructure characterization of apple and pear fruit tissue in X-ray micro-computed tomography (CT) images.

A novel self-supervised method for in-field occluded apple ripeness determination →

The full view of the apples in the orchard is often obscured by leaves and trunks, making it challenging to accurately determine their ripeness, whilst it is an essential yet difficult task for apple-harvesting robots. Within this context, we propose a novel method to address two critical challenges: ripeness determination and in-field occlusion. The proposed method is trained in a self-supervised manner on a dataset consisting of less than 1% labelled images and the rest of unlabelled images. It is made up of three key parts: a reconstructor, a feature extractor, and a predictor. The reconstructor is designed to reconstruct the missing parts of occluded apples. The feature extractor is introduced to learn ripeness-related features from the vast number of unlabelled images. Unlike the previous approaches classifying the fruit ripeness into several discrete categories, the predictor uses the learned features to generate a continuous ripeness score in the range between 0.0 and 1.0, thus eliminating the need to subjectively pre-define ripeness stages and offering end-users the flexibility to make their own decisions. Experimental results comparing our method to another method with different settings show that our method achieves the best Structural Similarity Index Measure (SSIM) of 0.75 and the second-best Peak-Signal-to-Noise Ratio (PSNR) of 25.36 for reconstructing missing apple parts, whilst using the fewest 86.3M parameters. Besides, our method outperforms 15 other self-supervised methods and even a supervised method in the ripeness score prediction, with the smallest score 0.0127 for fully unripe and the highest score 0.8933 for fully ripe apples. The results demonstrate the potential of our method to be incorporated with in-field robotic systems, enabling them to assess ripeness for selective harvesting effectively. It is helpful to monitor the overall ripeness of large orchards digitally, aid the decision-making processes and advance the goals of smart and precision agriculture.

Why it matches plant phenotyping methodsリンゴの遮蔽画像から連続的な成熟度スコアを推定する自己教師あり画像解析法を開発し、再構成性能と成熟度予測性能を比較検証しているため、植物フェノタイピング手法が中心である。

abstractwe propose a novel method to address two critical challenges: ripeness determination and in-field occlusion

ECNNA: Enhanced Convolutional Neural Network with Attention Mechanism for Plant Leaf Disease Classification →

Recently, the classification of plant leaf diseases has become a critical research area to improve the agricultural productivity.Early and accurate identification of diseases is needed to prevent the disease transmission and reduce the crop losses.Deep learning approaches enables to learn complex meaningful patterns within the various leaves.In the paper, enhanced convolutional neural network with attention mechanism (ECNNA) model is proposed to extract the discriminative features and classify the types of leaf diseases.The ECNNA comprises of prior feature extraction with convolution layers and maximum pooling layers, feature enhancement with attention mechanism, classification using SoftMax classifier.Additionally, the model utilizes data augmentation techniques to increase dataset diversity and improve generalization ability of model.The attention layer is incorporated in convolutional neural network to improve the performance of system that increase crop yield and quality.The system classifies the various diseases categories.Experimental results demonstrate that the proposed ECNNA model achieves the classification accuracy of 99.99% on Potato dataset, 97.27% on Corn dataset, 97.95% on Apple dataset, 99.14% on Grape dataset, 99.62% on Peach dataset, 99.31% on six classes dataset, and 98.90% on eight classes dataset.The classification results are also compared with previous studies, indicating the higher classification rate.This research contributes to the early detection and diagnosis of plant leaf disease for supporting sustainable agriculture and food security.

Why it matches plant phenotyping methods植物葉の画像から病害状態を分類する深層学習手法を開発し、複数データセットで性能比較・評価しており、病害表現型の取得・判定が中心です。

abstractenhanced convolutional neural network with attention mechanism (ECNNA) model is proposed to extract the discriminative features and classify the types of leaf diseases.

Green Symphony: Deep Learning for Crop Health Assessment →

Abstract: The agricultural sector holds paramount importance in our economy, impacting our daily lives significantly. Effective management of agricultural resources is crucial for ensuring profitability in crop production. However, farmers often lack expertise in identifying and managing plant leaf diseases, leading to reduced yields. Detecting and classifying leaf diseases is pivotal for maximizing agricultural productivity. Utilizing Convolutional Neural Networks (CNNs) offers a promising solution for automated leaf disease detection and classification. This research focuses on detecting diseases in key crops such as apple, grape, corn, potato, and tomato plants. By leveraging deep CNN models, this study aims to enhance disease monitoring in large crop fields, enabling prompt identification of disease symptoms and facilitating timely intervention. Such advancements in plant leaf disease detection have broad applications in biological research and agricultural institutes, offering immense potential to optimize crop health management and maximize yields. Comparing the proposed deep CNN model with established transfer learning approaches like VGG16 underscores the significance of this research endeavor in addressing the critical need for efficient disease detection and management in agriculture..

Why it matches plant phenotyping methods植物葉の病徴を画像から検出・分類するCNN手法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、表現型計測手法として採用。

abstractUtilizing Convolutional Neural Networks (CNNs) offers a promising solution for automated leaf disease detection and classification.

Macrostructure of Malus Leaves and Its Taxonomic Significance. →

Leaves are the most ubiquitous plant organs, whose macrostructures exhibit close correlations with environmental factors while simultaneously reflecting inherent genetic and evolutionary patterns. These characteristics render them highly significant for plant taxonomy, ecology, and related disciplines. Therefore, this study presents the first comprehensive evaluation of Malus leaf macrostructures for infraspecific classification. By establishing a trait-screening system, we conducted a numerical taxonomic analysis of leaf phenotypic variation across 73 Malus germplasm (34 species and 39 cultivars). Through ancestor-inclined distribution characteristic analysis, we investigated phylogenetic relationships at both the genus level and infraspecific ranks within Malus . A total of 21 leaf phenotypic traits were selected from 50 candidate traits based on the following criteria: high diversity, abundance, and evenness (D ≥ 0.50, H ≥ 0.80, and E ≥ 0.60); significant intraspecific uniformity and interspecific distinctness (CV¯ ≤ 10% and CV ≥ 15%). Notably, the selected traits with low intraspecific variability (CV¯ ≤ 10%) exhibit environmental robustness, likely reflecting low phenotypic plasticity of these specific traits under varying conditions. This stability enhances their taxonomic utility. It was found that the highest ancestor-inclined distribution probability reached 90% for 10 traceable cultivars, demonstrating reliable breeding lines. Based on morphological evidence, there was a highly significant correlation between the evolutionary orders of (Sect. Docyniopsis → Sect. Sorbomalus → Sect. Malus ) and group/sub-groups (B 1 → B 2 → A). This study demonstrates that phenotypic variation in leaf macrostructures can effectively explore the affinities among Malus germplasm, exhibiting taxonomic significance at the infraspecific level, thereby providing references for variety selection. However, hybrid offspring may exhibit mixed parental characteristics, leading to blurred species boundaries. And convergent evolution may create false homologies, potentially misleading morphology-based taxonomic inferences. The inferred taxonomic relationships present certain limitations that warrant further investigation.

Why it matches plant phenotyping methodsMalus葉の形態形質を体系的に選抜・評価するtrait-screening systemを構築し、50候補から21形質を定量基準で選定して分類へ適用しており、表現型取得・選抜手法が研究の中心である。

abstractBy establishing a trait-screening system, we conducted a numerical taxonomic analysis of leaf phenotypic variation across 73 Malus germplasm (34 species and 39 cultivars).

Detection of Plant Leaf Diseases using Machine Learning Techniques and CNN →

Agriculture is critical for guaranteeing food security, in turn maintaining the economic stability of a country. Diseases in plant leaves are considerable threat to plant health and productivity. Pathology expert perform manual identification of diseases which is time-consuming, labor-intensive and delayed detection can lead to serious crop loss. Early and accurate detection of diseases has to be done prefer ably. It is challenging to diagnosis diseases as leaves exhibit almost similar color, texture features. The focus is to identify and categorize plant leaf diseases with respect to five plant species apple, corn, grape, potato and tomato taken from PlantVillage dataset. These species are considered with over 25 classes that include healthy and diseased classes. The experiments are conducted by image processing techniques and Machine Learning (ML) models such as Decision Tree (DT), K-Nearest Neighbors (KNN), Random Forest (RF), Support Vector Machine (SVM), and Convolutional Neural Networks (CNN). All the above techniques are considered for comparison by performance evaluation metrics: accuracy, precision, recall, and F1-score. CNN outperforms here in plant disease detection by obtaining an accuracy of 95.37%, 93.92%, 98.00%, 97.10% and 94.88% with respect to apple, corn, grapes, potato and tomato crop leaves respectively.

Why it matches plant phenotyping methods植物葉の画像から病害状態を分類する機械学習・CNN手法を比較評価しており、植物病害表現型の取得・推定が中心です。

abstractThe focus is to identify and categorize plant leaf diseases with respect to five plant species apple, corn, grape, potato and tomato taken from PlantVillage dataset.

Nondestructive detection of apple watercore disease content based on 3D watercore model →

Current cultivation and research on Watercore apples lack precise evaluation methods and non-destructive detection techniques for Watercore content. In response, this study exploits the intrinsic distribution characteristics of Watercore and utilizes a RIFE interpolation-based feature slice stacking method to reconstruct a 3D model of individual Watercore—a task unattainable using conventional approaches. Employing the complete 3D Watercore model as a reference, the study further integrates near-infrared spectroscopy with the GAF-ConvNeXt algorithm to achieve five-class non-destructive detection of Watercore. Experimental results demonstrate that the MIoU between the RIFE-interpolated features and the original Watercore features attains a value of 0.826, thereby indicating high reliability. The reconstructed 3D models typically exhibit a central void, multiple uniformly distributed independent pillar-like structures along the periphery, and a greater volume in the upper half relative to the lower half. Furthermore, the five-class detection accuracy achieved using the GAF-ConvNeXt algorithm attains 98.10 %, thereby offering a more precise and scientifically robust method for the non-destructive evaluation of Watercore content in apples.

Why it matches plant phenotyping methodsリンゴのWatercore内容を対象に、3Dモデル再構成と近赤外分光による非破壊検出手法を開発・評価しており、植物状態の取得が研究の中心である。

abstractthis study exploits the intrinsic distribution characteristics of Watercore and utilizes a RIFE interpolation-based feature slice stacking method to reconstruct a 3D model of individual Watercore

Location-guided lesions representation learning via image generation for assessing plant leaf diseases severity →

Accurate assessment of plant leaf disease severity is crucial for implementing precision pesticide application, which in turn significantly enhances crop yields. Previous methods primarily rely on global perceptual learning, often leading to the misidentification of non-lesion regions as lesions within complex backgrounds, thereby compromising model accuracy. To address the challenge of background interference, we propose a location-guided lesion representation learning method (LLRL) based on image generation to assess the severity of plant leaf diseases. Our approach comprises three key parts: the image generation network (IG-Net), the location-guided lesion representation learning network (LGR-Net), and the hierarchical lesion fusion assessment network (HLFA-Net). IG-Net is designed to construct paired images necessary for LGR-Net by utilizing a diffusion model to generate diseased leaves from healthy ones. First, the LGR-Net facilitates the network's focus on the lesion area by contrasting paired images: healthy and diseased leaves, obtaining a pre-trained dual-branch feature encoder (DBF-Enc) that incorporates lesion-specific prior knowledge, providing focused visual features for HLFA-Net. Second, the HLFA-Net, which shares and freezes the DBF-Enc, further fuses and optimizes the features extracted by DBF-Enc, culminating in a precise classification of disease severity. In addition, we construct an image dataset containing three plant leaf diseases from apple, potato, and tomato plants, with a total of 12,098 photos, to evaluate our approach. Finally, experimental results demonstrate that our method outperforms existing classification models, with at least an improvement of 1 ​% in accuracy for severity assessment, underscoring the efficacy of the LLRL method in accurately identifying the severity of plant leaf diseases. Our code and dataset are available at http://llrl.samlab.cn/.

Why it matches plant phenotyping methods植物葉の病害重症度という状態を画像から推定する手法を開発し、データセットで評価しており、フェノタイピング手法が中心です。

abstractwe propose a location-guided lesion representation learning method (LLRL) based on image generation to assess the severity of plant leaf diseases.

Segmentation and classification of plant leaf disease using advanced deep learning approach and ensemble classifier →

An essential component of maintaining global food production is plants. On other hand, a number of plant diseases can threaten agricultural output and cause large losses if left unchecked. Agricultural specialists and botanists physically track plant diseases in a labor-intensive, error-prone manner using a conventional method. AI can give evaluations that are quicker and more accurate than those made using conventional approaches by automating the identification and analysis of diseases. This technical development presents a viable way to lessen crop losses and lessen the severity of infections. As a result, we describe an ensemble machine learning strategy for plant disease classification in this study that is enabled by deep learning. Data augmentation is done in the first part of the study, and in the second step, we provide a modified Mask R-CNN model for plant leaf segmentation. Afterwards, a model to extract the deep features based on CNN is shown. Lastly, the ensemble classifier is built using support vector machine classifier (SVM), random forest (RF), and decision tree (DT) with the aid of majority voting. The suggested method's effectiveness is tested on plant village, apple, maize, and rice, yielding overall accuracy values of 99.45%, 96.30%, 96.85%, and 98.25%, in that order.

Why it matches plant phenotyping methods植物葉の病害状態を画像からセグメンテーション・分類する深層学習手法の開発と精度評価が中心であり、植物フェノタイピング手法に該当する。

abstractwe provide a modified Mask R-CNN model for plant leaf segmentation

Unveiling the fingerprint of apple browning: A Vis/NIR-metaheuristic approach for rapid polyphenol oxidase and peroxidases activities detection in red delicious apples →

As a climacteric fruit, apple fruit quality during storage is influenced by the activity of two browning-related enzymes, polyphenol oxidase (PPO) and peroxidase (POD). Therefore, to evaluate the enzymatic activity of Red Delicious apples, the content of PPO and POD was measured using destructive chemical methods and used as the response for visible/near-infrared (Vis/NIR) spectroscopy. Different variable selection algorithms were implemented in combination with two machine learning algorithms of support vector machine (SVM) and decision tree (DT), to identify the effective wavelengths from the whole spectral data. DT-FOA (forest optimization algorithm) algorithm outperformed other methods in terms of minimum number of effective wavelengths (EWs), minimum execution time, and maximum correlation. Multiple linear regression (MLR), partial least squares regression (PLSR), and artificial neural network (ANN) were applied to predict enzymatic activities. The selection of the optimum predictive model was mainly based on criteria such as the coefficient of determination (R²), root mean square error (RMSE), the ratio of prediction to deviation (RPD) of the validation set. ANN outperformed the MLR and PLSR in terms of the highest R² (0.96 and 0.99) and RPD (4.87 and 6.96) in test phase of DT-FOA, for PPO and POD, respectively. However, all the model gave reliable results being the R² above 0.92 and 0.93, and RPD above 5.36 and 5.31 for MLR and PLSR in test phase of DT-FOA, for PPO and POD respectively. The combination of Vis/NIR spectroscopy, regression algorithm and variable selection led to a tool for evaluating Red Delicious apple fruit.

Why it matches plant phenotyping methodsリンゴ果実の褐変関連酵素活性という植物器官の状態を、Vis/NIR分光と波長選択・回帰モデルで非破壊推定する手法が研究の中心であり、モデル性能の検証も行っている。

abstractDifferent variable selection algorithms were implemented in combination with two machine learning algorithms of support vector machine (SVM) and decision tree (DT), to identify the effective wavelengths from the whole spectral data.

Explicit dynamics simulation study to determine the damage patterns of apples (Red Fuji) under impact loading →

Collisions between apples and mechanical structures are inevitable during the harvest of fresh-market fruit. Even moderate impacts can lead to internal damage, significantly affecting the shelf life and quality of the fruit. In recent years, technologies such as spatial frequency domain imaging and structured light reflection imaging have made progress in early damage detection. However, considering both economic feasibility and practical applicability, there remains a need to explore accurate and quantitative methods for damage assessment. To address this issue, this study constructed a 3D model of ‘Red Fuji’ apples through physical experiments and reverse engineering. Explicit dynamic simulations based on the finite element method were conducted to evaluate the mechanical damage under various impact conditions. A total of 125 simulation scenarios were designed by combining five impact heights, five impact angles, and five contact materials. Key data and visual representations of stress evolution were obtained from the simulations. Results indicated that the highest damage susceptibility occurred when apples impacted steel at a height of 20 cm and an angle of 135°, while the lowest damage susceptibility was observed when impacting polyvinyl chloride at a height of 15 cm and an angle of 90°. Furthermore, a response surface methodology was employed to analyze the quantitative values of damage susceptibility. The maximum discrepancies between experimental and simulated results in terms of damage depth, area, and volume were 0.90 cm, 2.94 cm², and 11.85 cm³, respectively. The prediction error of damage susceptibility ranged from 0.883 to 11.3 %. The consistency of the damage patterns further validates that the finite element model can effectively simulate apple damage under specific impact scenarios. This study provides insights for reducing mechanical damage during harvesting.

Why it matches plant phenotyping methodsリンゴ果実の損傷状態を有限要素シミュレーションで定量評価し、実験結果と損傷深さ・面積・体積および予測誤差を比較検証しているため、植物器官の状態推定手法が研究の中心である。

abstractExplicit dynamic simulations based on the finite element method were conducted to evaluate the mechanical damage under various impact conditions.

Leveraging Lesion Segmentation Masks to Validate CNN Focus in Apple Disease Classification using Explainable AI →

This paper presents a procedure to explain the focus of the Convolutional Neural Networks (CNNs) for classifying apple diseases. The goal of this work is to promote more transparency and trust in CNN-based diagnostic tools by using Explainable AI (XAI) methods -here Grad-CAM (Gradient-weighted Class Activation Mapping) in the agricultural setting. The main concept in the proposed pipeline is to use apple leaf images as well as the manually created lesion segmentation masks. A pre-trained CNN is used for disease classification, where the last two-weighted layers are employed to extract the significantly enriched features, and then Grad-CAM is used to output the heatmap to highlight the informative parts for the decision. One of the main contributions of this study is to quantitatively compare these Grad-CAM heatmaps with the ground truth labels (lesion masks) in terms of Intersection over Union (IoU) score. This test gives us a way to quantitatively evaluate if the CNN is learning from real disease symptoms. By making decisions about the model dependent on pathological features, this approach intends to provide the application of CNNs for apple disease classification with much valuable confidence and reliability.

Why it matches plant phenotyping methodsリンゴ葉の病徴（病斑）を画像から扱い、Grad-CAMの病斑適合性をIoUで定量検証する手法研究であり、疾病状態の推定・検証が中心である。

abstractOne of the main contributions of this study is to quantitatively compare these Grad-CAM heatmaps with the ground truth labels (lesion masks) in terms of Intersection over Union (IoU) score.

FruitNeRF++: A Generalized Multi-Fruit Counting Method Utilizing Contrastive Learning and Neural Radiance Fields →

FruitNeRF++: A Generalized Multi-Fruit Counting Method Utilizing Contrastive Learning and Neural Radiance Fields We introduce FruitNeRF++, a novel fruit-counting approach that combines contrastive learning with neural radiance fields to count fruits from unstructured input photographs of orchards. Our work is based on FruitNeRF [6], which employs a neural semantic field combined with a fruit-specific clusteringapproach. The requirement for adaptation for each fruit type limits the applicability of the method, and makes it difficult to use in practice. To lift this limitation, we design a shape-agnostic multi-fruit counting framework, that complements the RGB and semantic data with instance masks predicted by a vision foundation model. The masks are used to encode the identity of each fruit as instance embeddings into a neural instance field. By volumetrically sampling the neural fields, we extract apoint cloud embedded with the instance features, which can be clustered in a fruit-agnostic manner to obtain the fruit count. We evaluate our approach using a synthetic dataset containing apples, plums, lemons, pears, peaches, and mangoes, as well as a real-world benchmark apple dataset. Our results demonstrate that FruitNeRF++ is easier to control and compares favorably to other state-of-the-art methods.

Why it matches plant phenotyping methods果実を対象とした画像ベースの汎用カウント手法を開発し、合成および実データで評価しているため、植物形質（果実数）の取得・推定が研究の中心です。

abstractWe introduce FruitNeRF++, a novel fruit-counting approach that combines contrastive learning with neural radiance fields to count fruits from unstructured input photographs of orchards.

Location-guided lesions representation learning via image generation for assessing plant leaf diseases severity. →

Accurate assessment of plant leaf disease severity is crucial for implementing precision pesticide application, which in turn significantly enhances crop yields. Previous methods primarily rely on global perceptual learning, often leading to the misidentification of non-lesion regions as lesions within complex backgrounds, thereby compromising model accuracy. To address the challenge of background interference, we propose a location-guided lesion representation learning method (LLRL) based on image generation to assess the severity of plant leaf diseases. Our approach comprises three key parts: the image generation network (IG-Net), the location-guided lesion representation learning network (LGR-Net), and the hierarchical lesion fusion assessment network (HLFA-Net). IG-Net is designed to construct paired images necessary for LGR-Net by utilizing a diffusion model to generate diseased leaves from healthy ones. First, the LGR-Net facilitates the network's focus on the lesion area by contrasting paired images: healthy and diseased leaves, obtaining a pre-trained dual-branch feature encoder (DBF-Enc) that incorporates lesion-specific prior knowledge, providing focused visual features for HLFA-Net. Second, the HLFA-Net, which shares and freezes the DBF-Enc, further fuses and optimizes the features extracted by DBF-Enc, culminating in a precise classification of disease severity. In addition, we construct an image dataset containing three plant leaf diseases from apple, potato, and tomato plants, with a total of 12,098 photos, to evaluate our approach. Finally, experimental results demonstrate that our method outperforms existing classification models, with at least an improvement of 1 ​% in accuracy for severity assessment, underscoring the efficacy of the LLRL method in accurately identifying the severity of plant leaf diseases. Our code and dataset are available at http://llrl.samlab.cn/.

Why it matches plant phenotyping methods植物葉の病害重症度という可視的な状態を画像から推定する手法を開発・評価しており、病害フェノタイピングが研究の中心です。

abstractwe propose a location-guided lesion representation learning method (LLRL) based on image generation to assess the severity of plant leaf diseases.

Band-specific segmented extinction correction enhances apple soluble solids content prediction using VIS/NIR spectroscopy. →

Accurate prediction of apple soluble solids content (SSC) is essential for fruit quality evaluation. Aiming at the problem of spectral aberration caused by the variation of fruit diameter in the existing visible/near infrared spectroscopy (VIS/NIR) detection, a novel spectral correction strategy was proposed in this study. By systematically analysing the correlation law between light intensity attenuation and fruit size in different wavelength bands (600-1000 nm), it was found that the traditional single-parameter correction models (exponential function method, hyperbolic sine function method) had limitations of applicability in a wide spectral range. Based on this, this paper innovatively proposed the band-specific segmented extinction correction method, established the mapping relationship between spectral intervals and size compensation parameters, and realised the multi-band synergistic correction. Experiments showed that the extinction coefficient method based on the exponential function and the diameter transformation method based on the hyperbolic sine function exhibited significant results in spectral correction for specific bands, resulting in a 6 %-10 % improvement in modelling accuracy at global size. However, when we introduced the band-specific segmented extinction correction, the spectra were effectively corrected over the entire band range, the light intensity differences between samples of different sizes were significantly reduced, and the modelling accuracy at global size jumped by 15 %. Specifically, the partial least squares regression (PLSR) model had a coefficient of determination (R 2 ) of 0.90 and a root mean square error (RMSE) of 0.55, and the convolutional neural network (CNN) model had the R 2 of 0.95 and the RMSE of 0.44, after corrected for the band-specific segmented extinction. Finally, this paper set up additional validation experiments to test the calibration effect of the three methods, and the results showed that band-specific segmented extinction correction method improved the modelling effect of the model most significantly. Therefore, the band-specific segmented extinction correction method proposed could effectively reduce the effect of apple diameter on the transmission spectrum and further improve the apple SSC's prediction accuracy.

Why it matches plant phenotyping methodsリンゴ果実の可溶性固形分を推定するVIS/NIR分光法について、果径補正アルゴリズムを開発し、追加検証実験で性能を比較しており、植物形質取得法が研究の中心である。

abstracta novel spectral correction strategy was proposed in this study

Hyperspectral imaging and machine learning for quality assessment of apples with different bagging types. →

This study examined how different bagging types (unbagged, mesh-bagged, and paper-bagged) affect the internal and external quality of fruits. Spectral images of 307 apples were collected using a visible-near infrared hyperspectral imaging system, and differences in color, size, firmness, soluble solids content (SSC), and aroma were analyzed. The effectiveness of various algorithms for extracting effective wavelengths was compared. Developed apple quality inspection models using various machine learning algorithms, achieving the best-performing model, CARS-MLR, with all indicators having an R 2 p greater than 0.75. The results revealed significant differences in quality indicators based on bagging type, with paper-bagged apples showing the best overall quality-brighter color, moderate size and firmness, and rich aroma. This suggests that hyperspectral imaging, feature selection algorithms and machine learning methods, is an effective approach for non-destructive quality assessment of apples. This research offers valuable insights for assessing the quality of other fruits with different bagging types.

Why it matches plant phenotyping methodsリンゴ果実の色・サイズ・硬度・SSC・香気を非破壊推定するハイパースペクトル画像と機械学習モデルが研究の中心であり、波長選択アルゴリズムと検査モデルの性能も比較しているため。

abstractDeveloped apple quality inspection models using various machine learning algorithms, achieving the best-performing model, CARS-MLR, with all indicators having an R 2 p greater than 0.75.

Preliminary development of a new multispectral vision-based, automated apple grading system towards in-field fruit presorting →

While Machine vision technology has been widely implemented for fruit quality inspection at packing line facilities, but appropriate in-orchard pre-sorting technology has yet to be developed. This study represents a novel effort to leverage advanced real-time multispectral vision coupled with artificial intelligence to develop a new, automated apple grading system that inspects size, color, and surface defects simultaneously, towards in-orchard application. The system consists of a multispectral imaging chamber on a compact screw conveyor, which acquires five-band images from singulated apples traveling and rotating on the conveyor. Online experiments are conducted on different varieties of apples in diverse quality conditions at different conveyor speeds. A deep learning-based computer vision algorithm pipeline is developed to segment and track each apple on the conveyor while assessing its quality attributes (size, color, and surface defects) from different, multiple views, and grade the fruit based on full-surface quality information into three quality categories. The system will evolve into a fully integrated machine prototype for automated, in-orchard sorting.

Why it matches plant phenotyping methodsリンゴ果実のサイズ・色・表面欠陥という器官形質を、マルチスペクトル画像と深層学習で取得・評価する自動システムの開発が中心であり、単なる農業実験の routine 測定ではない。

abstractThis study represents a novel effort to leverage advanced real-time multispectral vision coupled with artificial intelligence to develop a new, automated apple grading system that inspects size, color, and surface defects simultaneously

AppleGrowthVision: A large-scale stereo dataset for phenological analysis, fruit detection, and 3D reconstruction in apple orchards →

Deep learning has transformed computer vision for precision agriculture, yet apple orchard monitoring remains limited by dataset constraints. The lack of diverse, realistic datasets and the difficulty of annotating dense, heterogeneous scenes. Existing datasets overlook different growth stages and stereo imagery, both essential for realistic 3D modeling of orchards and tasks like fruit localization, yield estimation, and structural analysis. To address these gaps, we present AppleGrowthVision, a large-scale dataset comprising two subsets. The first includes 9,317 high resolution stereo images collected from a farm in Brandenburg (Germany), covering six agriculturally validated growth stages over a full growth cycle. The second subset consists of 1,125 densely annotated images from the same farm in Brandenburg and one in Pillnitz (Germany), containing a total of 31,084 apple labels. AppleGrowthVision provides stereo-image data with agriculturally validated growth stages, enabling precise phenological analysis and 3D reconstructions. Extending MinneApple with our data improves YOLOv8 performance by 7.69 % in terms of F1-score, while adding it to MinneApple and MAD boosts Faster R-CNN F1-score by 31.06 %. Additionally, six BBCH stages were predicted with over 95 % accuracy using VGG16, ResNet152, DenseNet201, and MobileNetv2. AppleGrowthVision bridges the gap between agricultural science and computer vision, by enabling the development of robust models for fruit detection, growth modeling, and 3D analysis in precision agriculture. Future work includes improving annotation, enhancing 3D reconstruction, and extending multimodal analysis across all growth stages.

Why it matches plant phenotyping methodsリンゴの生育段階・果実・樹体構造を対象とする大規模ステレオ画像データセットを構築し、果実検出、フェノロジー分析、3D再構成モデルの評価に用いており、表現型取得・解析基盤が中心である。

titleAppleGrowthVision: A large-scale stereo dataset for phenological analysis, fruit detection, and 3D reconstruction in apple orchards

Deep Learning-Based Early Detection of Crop Diseases Using Leaf Image Analysis in Smart Agricultural Systems →

Early and accurate detection of crop diseases is critical for global food security and efficient agricultural management. Recent advances in deep learning, particularly convolutional neural networks (CNNs) and vision transformers (ViT), have demonstrated exceptional ability to recognize disease symptoms from leaf images. In this article, we present a comprehensive framework for plant disease detection that integrates state-of-the-art deep learning models into smart agriculture systems. We review publicly available datasets (e.g. the PlantVillage dataset with 54,306 leaf images across 14 crop species and 26 disease classes), and discuss data preprocessing and augmentation techniques. We then detail various model architectures: traditional CNNs (e.g. ResNet, MobileNet), efficient CNN variants, ViT-based models, and hybrid CNN–ViT architectures (e.g. FOTCA, AppViT). Our proposed models leverage transfer learning and attention mechanisms to improve accuracy. We describe an experimental setup using multiple leaf-image datasets (tomato, potato, apple, cassava, wheat) and report hypothetical results: for example, our hybrid model achieves ≈99.7% accuracy on PlantVillage and 98–99% on tomato/potato datasets. We include precision, recall, F1 metrics and confusion matrices to analyze performance. Integration into smart farming is discussed: IoT sensors and mobile devices capture leaf images, which are processed by on-device or cloud CNN/ViT models to alert farmers in real time, the depthwise separable convolution block, and the ViT encoding block, respectively. We compare results across models and examine the trade-offs between model complexity and accuracy. Our findings confirm that hybrid CNN–ViT architectures yield the best performance, while lightweight models (e.g. MobileViT, AppViT) enable on-device inference.

Why it matches plant phenotyping methods葉画像から植物病害症状を推定する深層学習手法の開発・比較・評価が中心であり、植物の病害状態を直接推定する画像ベース表現型解析に該当する。

abstractwe present a comprehensive framework for plant disease detection that integrates state-of-the-art deep learning models into smart agriculture systems.

Apple Yield Estimation Method Based on CBAM-ECA-Deeplabv3+ Image Segmentation and Multi-Source Feature Fusion. →

Apple yield estimation is a critical task in precision agriculture, challenged by complex tree canopy structures, growth stage variability, and orchard heterogeneity. In this study, we apply multi-source feature fusion by combining vegetation indices from UAV remote sensing imagery, structural feature ratios from ground-based fruit tree images, and leaf chlorophyll content (SPAD) to improve apple yield estimation accuracy. The DeepLabv3+ network, optimized with Convolutional Block Attention Module (CBAM) and Efficient Channel Attention (ECA), improved fruit tree image segmentation accuracy. Four structural feature ratios were extracted, visible-light and multispectral vegetation indices were calculated, and feature selection was performed using Pearson's correlation coefficient analysis. Yield estimation models were constructed using k-nearest neighbors (KNN), partial least squares (PLS), random forest (RF), and support vector machine (SVM) algorithms under both single feature sets and combined feature sets (including vegetation indices, structural feature ratios, SPAD, vegetation indices + SPAD, vegetation indices + structural feature ratios, structural feature ratios + SPAD, and the combination of all three). The optimized CBAM-ECA-DeepLabv3+ model achieved a mean Intersection over Union (mIoU) of 0.89, an 8% improvement over the baseline DeepLabv3+, and outperformed U2Net and PSPNet. The SVM model based on multi-source feature fusion achieved the highest apple yield estimation accuracy in small-scale orchard sample plots (R 2 = 0.942, RMSE = 12.980 kg). This study establishes a reliable framework for precise fruit tree image segmentation and early yield estimation, advancing precision agriculture applications.

Why it matches plant phenotyping methods画像分割による果樹構造形質の抽出と、UAV・地上画像・SPADの融合によるリンゴ収量推定が中心であり、手法の精度比較・検証も明示されているため。

abstractThe DeepLabv3+ network, optimized with Convolutional Block Attention Module (CBAM) and Efficient Channel Attention (ECA), improved fruit tree image segmentation accuracy.

Integrating Local Texture Capturing Mechanisms With Convolutional Neural Networks For Enhanced Multi‑Class Classification of Plant Leaf Diseases →

Plant diseases significantly affect agricultural productivity by reducing both the quality and quantity of crops. The necessity for automated image‑based solutions stems from the labor‑intensive and subjectively error‑prone nature of traditional inspection methods performed by farmers or agricultural specialists. To maintain sustainable agriculture and prevent the spread of infections, the detection of plant leaf diseases should be performed early and accurately. Early identification of infections can also significantly reduce yield losses and minimize the excessive use of pesticides. Since leaf diseases frequently manifest as uneven texture patterns, spots, or distortions on the leaf surface, local texture capturing mechanisms have proven to be remarkably effective among many computational approaches. This study proposes a novel Deep Convolutional Neural Network (DCNN) to extract high‑level hidden feature representations from leaf images. To enhance performance, the deep features are combined with traditional handcrafted texture features known as the Uniform Local Binary Pattern (uLBP). The proposed model was trained and tested using three well‑known publicly available datasets: Apple Leaf, Tomato Leaf, and Grape Leaf. The model achieved test accuracies of 96%, 91%, and 96% on these datasets, respectively. The experimental results demonstrate that the proposed approach is an effective and practical method for early diagnosis of plant diseases. This system has potential for real‑world application by farmers and agricultural experts to support disease management and contribute to the development of more resilient crops and a sustainable agricultural industry.

Why it matches plant phenotyping methods植物葉画像から病害状態を推定する画像解析手法を提案・評価しており、病害表現型の取得・分類が研究の中心であるため。

abstractThis study proposes a novel Deep Convolutional Neural Network (DCNN) to extract high‑level hidden feature representations from leaf images.

Artificial Neural Networks for Image Processing in Precision Agriculture: A Systematic Literature Review on Mango, Apple, Lemon, and Coffee Crops →

Precision agriculture is an approach that uses information technologies to improve and optimize agricultural production. It is based on the collection and analysis of agricultural data to support decision making in agricultural processes. In recent years, Artificial Neural Networks (ANNs) have demonstrated significant benefits in addressing precision agriculture needs, such as pest detection, disease classification, crop state assessment, and soil quality evaluation. This article aims to perform a systematic literature review on how ANNs with an emphasis on image processing can assess if fruits such as mango, apple, lemon, and coffee are ready for harvest. These specific crops were selected due to their diversity in color and size, providing a representative sample for analyzing the most commonly employed ANN methods in agriculture, especially for fruit ripening, damage, pest detection, and harvest prediction. This review identifies Convolutional Neural Networks (CNNs), including commonly employed architectures such as VGG16 and ResNet50, as highly effective, achieving accuracies ranging between 83% and 99%. Additionally, it discusses the integration of hardware and software, image preprocessing methods, and evaluation metrics commonly employed. The results reveal the notable underuse of vegetation indices and infrared imaging techniques for detailed fruit quality assessment, indicating valuable opportunities for future research.

Why it matches plant phenotyping methods果実の成熟・損傷・病害などを画像処理とANNで評価する方法を体系的にレビューしており、植物表現型取得・推定手法が中心である。

abstractThis article aims to perform a systematic literature review on how ANNs with an emphasis on image processing can assess if fruits such as mango, apple, lemon, and coffee are ready for harvest.

Plant Leaf Disease Detection System Using CNN →

Abstract: Fruit and vegetable crops experience diminished agricultural production output because of pests along with diseases that rank as major factors worldwide. The correct identification of these issues becomes vital because delayed detection results in decreased quantity and quality of yields which then causes problems for food supply networks and regional economic stability. Farmers traditionally monitor plant diseases through individual observation assisted by expert consultations because they search for clear indicators of leaf damage including discolorations or spotted lesions or deteriors on leaves. The approach fails to meet standards because it often produces unsuitable results and unreliable human involvement. The proposed deep learning-based Disease Recognition Model employs Convolutional Neural Networks (CNNs) for processing leaf disease diagnosis within apple and corn and tomato and potato crops. The system enables automatic disease detection of leaves through image processing which delivers precise results. The training data consists of multiple leaf images which come from healthy subjects and disease-infected samples enabling precise identification of various diseases. The tool aims to become an affordable solution that supports farmers and agronomists and policymakers for better crop health management and minimal chemical usage while ensuring sustainable farming practices . Key Words: The system employs key terms including Leaf disease detection, plant health monitoring, CNN classification, fruit and vegetable crops, automated diagnosis, early disease intervention, sustainable agriculture, precision farming.

Why it matches plant phenotyping methods葉画像から植物病害状態をCNNで推定する画像ベースの植物フェノタイピング手法が研究の中心であるため。

abstractThe proposed deep learning-based Disease Recognition Model employs Convolutional Neural Networks (CNNs) for processing leaf disease diagnosis within apple and corn and tomato and potato crops.

Branch segmentation and phenotype extraction of apple trees based on improved Laplace algorithm →

Phenotypic traits of crops reflect their physiological characteristics and provide a theoretical basis for predicting their growth. The 3D point cloud has a direct and accurate rendering ability, which has been widely used in phenotype extraction, especially with the help of accurate segmentation techniques. However, the inherent discrete nature of point clouds makes accurate organ segmentation an ongoing challenge in the field. In this study, we propose a tree phenotype acquisition method based on point cloud registration and skeleton segmentation. First, the Convex Hull-indexed Gaussian Mixture Model (CH-GMM) is employed to register the ground and aerial point cloud data. Then, a Laplace-multi-scale adaptive algorithm (LMSA) was proposed to obtain the crop skeleton structure, on the basis of which four phenotypic parameters, namely, plant height, crown width, branching number, and initial branching height, were extracted for fruit trees. In addition, the relationship between crown width and the number of branches was explored, where branches included initial, secondary, and tertiary branches. The results show that the proposed CH-GMM algorithm has a rotation error of less than 1.01°, a translation error of less than 10 mm, and a success rate of more than 95 %. The average precision, average recall, average F1 score, and average overall accuracy of the LMSA are 93.7 %, 96.2 %, 92.6 %, and 95.3 %, respectively. Finally, this study found a polynomial and exponential relationship between the number of bifurcations and crown size of fruit trees. The results of this study may provide new ideas for fruit tree phenotype acquisition and phenotype management.

Why it matches plant phenotyping methodsリンゴ樹の3D点群登録・骨格分割アルゴリズムを開発し、樹高や樹冠幅などの形態形質を抽出・精度評価しており、表現型取得手法が研究の中心である。

abstractwe propose a tree phenotype acquisition method based on point cloud registration and skeleton segmentation.

Plant Pathology Identification Using Digital Imaging →

Abstract: This study presents an innovative system for identifying crop diseases using a deep learning approach based on the Mobile Net architecture. Designed for efficiency and lightweight performance, Mobile Net enables accurate disease detection from leaf images while being highly suitable for deployment on mobile devices. The system incorporates a user-friendly graphical interface and a dedicated mobile application, allowing farmers to upload leaf images directly from their smartphones and receive instant diagnoses along with recommended treatments. Trained on the Plant Village dataset, the model is optimized for identifying diseases affecting five major crops: corn, apple, sugarcane, wheat, and grapes. By surpassing the limitations of traditional methods such as K-means clustering and SVM, the proposed system offers higher accuracy, faster processing, and real-time accessibility. This solution aims to minimize crop losses, improve agricultural productivity, and empower farmers with a portable and practical tool for effective crop management.

Why it matches plant phenotyping methods葉画像から作物病害を推定する深層学習システムを開発しており、植物の病害状態を直接評価する方法が中心である。

abstractThis study presents an innovative system for identifying crop diseases using a deep learning approach based on the Mobile Net architecture.

Apple phenotyping using deep learning and 3D depth analysis: An experimental study on fruitlet sizing during early development →

• Achieved high detection accuracy of apple fruitlets in complex orchard environments with rapid phenological changes. • Provided a dataset of videos and RGB-D images, featuring annotated apple fruitlets and manual caliper measurements during early development. • Developed a workflow for rapid in-field monitoring of flower corymbs and fruitlet sizing, validated through experimental trials. Current research in apple-growing focuses on collecting extensive biometric data to better understand physiological processes, improve orchard productivity and predict yields. In this context, fruit thinning has emerged as a key horticultural practice to enhance fruit size and quality while preventing alternate bearing. Despite the growing role of plant imaging technologies in agronomic management, fruitlet sizing remains challenging, particularly in early phenological stages. To address this challenge, we developed an RGB-D-based vision pipeline that combines YOLO models with depth information and relies on the statistical analysis of frame series to detect and cluster fruitlets into flower corymbs, providing both fruitlet counting and diameter estimates for each video acquisition. After obtaining an AP@0.5 and AP@[0.5:0.95] of respectively 0.894 and 0.77 in fruitlet detection, along with a precision of 0.881 and a recall of 0.846, our approach efficiently processed video frames, extracting the most reliable data for each labeled cluster. While the comparison of true positive estimates with calibrated caliper measurements showed a mean RMSE of 1.05 mm, challenges remain in achieving the correct fruitlet count, with a mean counting error of 0.63 fruitlets per video. Additionally, the proposed workflow retrieved the exact number of fruitlets as the ground truth in 56.4% of the videos, increasing to 75% when excluding those videos where the correct fruitlet count was never detected in any frame by the YOLO model. Despite these limitations, our results are promising, proposing a potential data acquisition tool without compromising the reliability of traditional practices. This approach could pave the way for future applications, including the evaluation of plant growth regulator trials and the development of predictive models for yield and productivity optimization.

Why it matches plant phenotyping methodsRGB-D画像と深度情報、YOLO、動画フレーム統計を組み合わせ、リンゴ果実の検出・計数・直径推定を行うワークフローを開発し、ノギス測定で検証しているため、表現型取得手法が中心である。

abstractProvided a dataset of videos and RGB-D images, featuring annotated apple fruitlets and manual caliper measurements during early development.

Detection of Heavy Metal Copper Stress in Apple Rootstocks Using Surface-Enhanced Raman Spectroscopy. →

Excessive use of copper (Cu) chemicals has led to soil contamination. This study utilized surface-enhanced Raman spectroscopy (SERS) to investigate the effects of 10 commonly encountered concentrations of Cu stress in orchards on apple rootstocks. Spectral preprocessing methods were employed to eliminate baseline drift and fluorescence background interference from the Raman spectra, while data augmentation techniques were incorporated to develop a one-dimensional stacked autoencoder convolutional neural network (1D-SAE-CNN) for classifying Cu stress levels, resulting in evaluation indices greater than 0.9. Scanning electron microscopy with energy dispersive spectroscopy (SEM-EDS) quantified Cu distribution in root, stem, and leaf tissues, while micro-Raman imaging visualized lignin, cellulose, and pigments under Cu stress. The results indicate that SERS combined with a deep learning model enables rapid and accurate differentiation of Cu stress levels in apple rootstocks in orchards, while SEM-EDS and micro-Raman imaging techniques reveal the migration effect of Cu 2+ within apple rootstock tissues and the ″low concentration promotion, high concentration inhibition″ effect of Cu on apple rootstock growth. Therefore, this approach showcases rapid and accurate detection of heavy metal Cu stress in apple rootstock tissues and has great potential for analyzing various types of heavy metal pollution in agricultural orchard ecosystems.

Why it matches plant phenotyping methodsSERSと深層学習を用いてリンゴ台木の銅ストレスレベルを分類する手法が研究の中心であり、植物のストレス状態を直接推定している。

abstractSERS combined with a deep learning model enables rapid and accurate differentiation of Cu stress levels in apple rootstocks

Innovative Machine Vision for Detecting Plant Disease Via Leaf Morphology and Recommending Fertilizers →

Plant diseases are highly influential on agricultural productivity and sustainability, hence the importance of effective and reliable predictive models. The paper covers prediction models. By using Apple leaf photos, the model will assist in predicting plant illness. The research focuses mostly on the quality of output. In addition to the improvement in output quality, this study addresses some of the challenges faced in previous models by incorporating the latest input features. This research proposes a model that will predict the disease using the VGG16 algorithm For accuracy, the model uses VIT to predict the disease stage. Following that, Random Forest Algorithm was used to select the most appropriate fertilizer for the plant based on the plant's disease stage. The framework proposed here attempts to help the farmer and the agricultural expert by giving an overall solution for early detection of disease stages, promoting sustainable agriculture. The research achieved an accuracy of 79% while training the VIT model and an accuracy of 98% with the VGG16 model.

Why it matches plant phenotyping methods葉画像から植物病害と病期を推定する画像解析モデルが研究の中心であり、植物の病害状態を直接評価するフェノタイピング手法に該当する。肥料推薦は付随的要素。

titleInnovative Machine Vision for Detecting Plant Disease Via Leaf Morphology and Recommending Fertilizers

CEFW-YOLO: A High-Precision Model for Plant Leaf Disease Detection in Natural Environments →

The accurate and rapid detection of apple leaf diseases is a critical component of precision management in apple orchards. The existing deep-learning-based detection algorithms for apple leaf diseases typically demand high computational resources, which limits their practical applicability in orchard environments. Furthermore, the detection of apple leaf diseases in natural settings faces significant challenges due to the diversity of disease types, the varied morphology of affected areas, and the influence of factors such as lighting variations, leaf occlusions, and differences in disease severity. To address the above challenges, we constructed an apple leaf disease detection (ALD) dataset, which was collected from real-world scenarios, and we applied data augmentation techniques, resulting in a total of 9808 images. Based on the ALD dataset, we proposed a lightweight YOLO11n-based detection network, named CEFW-YOLO, designed to tackle the current issues in apple leaf disease identification. First, we designed a novel channel-wise squeeze convolution (CWSConv), which employs channel compression and standard convolution to reduce computational resource consumption, enhance the detection of small objects, and improve the model’s adaptability to the morphological diversity of apple leaf diseases and complex backgrounds. Second, we developed an enhanced cross-channel attention (ECCAttention) module and integrated it into the C2PSA_ECCAttention module. By extracting global information, combining horizontal and vertical convolutions, and strengthening cross-channel interactions, this module enables the model to more accurately capture disease features on apple leaves, thereby enhancing detection accuracy and robustness. Additionally, we introduced a new fine-grained multi-level linear attention (FMLAttention) module, which utilizes multi-level asymmetric convolutions and linear attention mechanisms to improve the model’s ability to capture fine-grained features and local details critical for disease detection. Finally, we incorporated the Wise-IoU (WIoU) loss function, which enhances the model’s ability to differentiate overlapping targets across multiple scales. A comprehensive evaluation of CEFW-YOLO was conducted, comparing its performance against state-of-the-art (SOTA) models. CEFW-YOLO achieved a 20.6% reduction in computational complexity. Compared to the original YOLO11n, it improved detection precision by 3.7%, with the mAP@0.5 and mAP@0.5:0.95 increasing by 7.6% and 5.2%, respectively. Notably, CEFW-YOLO outperformed advanced SOTA algorithms in apple leaf disease detection, underscoring its practical application potential in real-world orchard scenarios.

Why it matches plant phenotyping methodsリンゴ葉の病害状態を画像から検出・推定する深層学習モデルとデータセットを開発し、既存モデルとの性能比較で技術的に評価しているため、植物フェノタイピング手法が中心である。

abstractwe constructed an apple leaf disease detection (ALD) dataset

A Comprehensive Review of Deep Learning in Computer Vision for Monitoring Apple Tree Growth and Fruit Production. →

The high nutritional and medicinal value of apples has contributed to their widespread cultivation worldwide. Unfavorable factors in the healthy growth of trees and extensive orchard work are threatening the profitability of apples. This study reviewed deep learning combined with computer vision for monitoring apple tree growth and fruit production processes in the past seven years. Three types of deep learning models were used for real-time target recognition tasks: detection models including You Only Look Once (YOLO) and faster region-based convolutional network (Faster R-CNN); classification models including Alex network (AlexNet) and residual network (ResNet); segmentation models including segmentation network (SegNet), and mask regional convolutional neural network (Mask R-CNN). These models have been successfully applied to detect pests and diseases (located on leaves, fruits, and trunks), organ growth (including fruits, apple blossoms, and branches), yield, and post-harvest fruit defects. This study introduced deep learning and computer vision methods, outlined in the current research on these methods for apple tree growth and fruit production. The advantages and disadvantages of deep learning were discussed, and the difficulties faced and future trends were summarized. It is believed that this research is important for the construction of smart apple orchards.

Why it matches plant phenotyping methodsリンゴ樹の生育、器官、収量、病害を画像・深層学習で評価する方法を中心にレビューしており、植物フェノタイピング手法レビューに該当する。

abstractThis study reviewed deep learning combined with computer vision for monitoring apple tree growth and fruit production processes in the past seven years.

AI Driven Crop Disease Prediction and Management System →

Agriculture is extremely important to human civilization, providing food and contributing to the economy. Plants are often susceptible to diseases and insects that have considerable challenges during production. Early detection of harvest diseases is important to minimize damage and reduce costs. While traditional methods do not provide real-time identification, foldable neuronal networks (CNNs) provide a solution by allowing for accurate detection and classification of leaf disease. This study focuses on identifying diseases in plants such as apples, grapes, corn, potatoes and tomatoes. The proposed deep CNN model is compared to a transfer learning approach, such as VGG16. AI-based systems analyze plant images to recognize diseases at the early stages and recommend management strategies, loss of harvests and improved yields. Such systems have applications in agriculture and biological research.

Why it matches plant phenotyping methods植物画像から葉の病害をCNNで検出・分類する手法の開発と比較が中心であり、病害状態という植物表現型を直接推定しているため。

abstractfoldable neuronal networks (CNNs) provide a solution by allowing for accurate detection and classification of leaf disease

An improved DeepLabV3+ based approach for disease spot segmentation on apple leaves →

This study presents an improved DeepLabV3+ model named AS-DeepLabV3+, specifically designed for segmenting disease spots on apple leaves. AS-DeepLabV3+ addresses critical challenges such as blurry spot edges, the small proportion of spot pixels, and significant variation in spot shapes. First, MobileNetV2 is employed as a lightweight backbone, reducing model complexity. Second, multiple attention mechanisms—Coordinate Attention, ECA Attention, CBAM, and Triplet Attention—are integrated into a unified Multi-Attention module, enhancing feature representation. Third, a dynamic Atrous Spatial Pyramid Pooling (ASPP) module is introduced to effectively capture multi-scale features. Lastly, dense connectivity is utilized in the decoder to improve feature reuse and detail recovery. The model was trained and validated on a dataset of 6,400 apple leaf images collected under natural lighting conditions. Experimental results demonstrate that our proposed model achieves a mean Intersection over Union (mIoU) of 98.00 %, a mean Pixel Accuracy (mPA) of 98.95 %, and a precision of 98.45 %, significantly outperforming existing models, including the original DeepLabV3+, SegNet, BiSeNet, PSPNet, and U-Net. Furthermore, the model has been integrated into a WeChat Mini Program to offer efficient and reliable disease detection services for agricultural practitioners.

Why it matches plant phenotyping methodsリンゴ葉の病斑を画像から分割・検出する手法の開発と検証が研究の中心であり、植物の病害状態を直接推定しているため。

abstractThis study presents an improved DeepLabV3+ model named AS-DeepLabV3+, specifically designed for segmenting disease spots on apple leaves.

Leveraging the internet of things and optimized deep residual networks for improved foliar disease detection in apple orchards. →

Plant diseases significantly threaten food security by reducing the quantity and quality of agricultural products. This paper presents a deep learning approach for classifying foliar diseases in apple plants using the Tunicate Swarm Sine Cosine Algorithm-based Deep Residual Network (TSSCA-based DRN). Cluster heads in simulated Internet of Things (IoT) networks are selected by Fractional Lion Optimization (FLION), and images are pre-processed with a Gaussian filter and segmented using the DeepJoint model. The TSSCA, combining the Tunicate Swarm Algorithm (TSA) and Sine Cosine Algorithm (SCA), enhances the classifier's effectiveness. Moreover, Plant Pathology 2020 - FGVC7 dataset is used in this work. This dataset is designed for the classification of foliar diseases in apple trees. The TSSCA-based DRN outperforms other methods, achieving 97% accuracy, 94.666% specificity, 96.888% sensitivity, and 0.0442J maximal energy, with significant improvements over existing approaches. Additionally, the proposed model demonstrates superior accuracy, outperforming other methods by 8.97%, 6.58%, 2.07%, 1.71%, 1.14%, 1.07%, 0.93%, and 0.64% over Multidimensional Feature Compensation Residual neural network (MDFC - ResNet), Convolutional Neural Network (CNN), Multi-Context Fusion Network (MCFN), Advanced Segmented Dimension Extraction (ASDE), and DRN, fuzzy deep convolutional neural network (FCDCNN), ResNet9-SE, Capsule Neural Network (CapsNet), IoT-based scrutinizing model, and Multi-Model Fusion Network (MMF-Net).

Why it matches plant phenotyping methodsリンゴ葉の病害状態を画像から分類する深層学習手法の開発と比較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。

abstractThis paper presents a deep learning approach for classifying foliar diseases in apple plants using the Tunicate Swarm Sine Cosine Algorithm-based Deep Residual Network (TSSCA-based DRN).

An enhanced lightweight model for apple leaf disease detection in complex orchard environments. →

Automated detection of apple leaf diseases is crucial for predicting and preventing losses and for enhancing apple yields. However, in complex natural environments, factors such as light variations, shading from branches and leaves, and overlapping disease spots often result in reduced accuracy in detecting apple diseases. To address the challenges of detecting small-target diseases on apple leaves in complex backgrounds and difficulty in mobile deployment, we propose an enhanced lightweight model, ELM-YOLOv8n.To mitigate the high consumption of computational resources in real-time deployment of existing models, we integrate the Fasternet Block into the C2f of the backbone network and neck network, effectively reducing the parameter count and the computational load of the model. In order to enhance the network's anti-interference ability in complex backgrounds and its capacity to differentiate between similar diseases, we incorporate an Efficient Multi-Scale Attention (EMA) within the deep structure of the network for in-depth feature extraction. Additionally, we design a detail-enhanced shared convolutional scaling detection head (DESCS-DH) to enable the model to effectively capture edge information of diseases and address issues such as poor performance in object detection across different scales. Finally, we employ the NWD loss function to replace the CIoU loss function, allowing the model to locate and identify small targets more accurately and further enhance its robustness, thereby facilitating rapid and precise identification of apple leaf diseases. Experimental results demonstrate ELM-YOLOv8n's effectiveness, achieving 94.0% of F1 value and 96.7% of mAP50 value-a significant improvement over YOLOv8n. Furthermore, the parameter count and computational load are reduced by 44.8% and 39.5%, respectively. The ELM-YOLOv8n model is better suited for deployment on mobile devices while maintaining high accuracy.

Why it matches plant phenotyping methodsリンゴ葉の病徴を画像から検出する軽量モデルを開発し、精度・計算量・モバイル展開性を評価しているため、植物病害状態の表現型取得手法が中心である。

abstractwe propose an enhanced lightweight model, ELM-YOLOv8n.

YO-AFD: an improved YOLOv8-based deep learning approach for rapid and accurate apple flower detection. →

The timely and accurate detection of apple flowers is crucial for assessing the growth status of fruit trees, predicting peak blooming dates, and early estimating apple yields. However, challenges such as variable lighting conditions, complex growth environments, occlusion of apple flowers, clustered flowers and significant morphological variations, impede precise detection. To overcome these challenges, an improved YO-AFD method based on YOLOv8 for apple flower detection was proposed. First, to enable adaptive focus on features across different scales, a new attention module, ISAT, which integrated the Inverted Residual Mobile Block (IRMB) with the Spatial and Channel Synergistic Attention (SCSA) module was designed. This module was then incorporated into the C2f module within the network's neck, forming the C2f-IS module, to enhance the model's ability to extract critical features and fuse features across scales. Additionally, to balance attention between simple and challenging targets, a regression loss function based on Focaler Intersection over Union (FIoU) was used for loss function calculation. Experimental results showed that the YO-AFD model accurately detected both simple and challenging apple flowers, including small, occluded, and morphologically diverse flowers. The YO-AFD model achieved an F1 score of 88.6%, mAP50 of 94.1%, and mAP50-95 of 55.3%, with a model size of 6.5 MB and an average detection speed of 5.3 ms per image. The proposed YO-AFD method outperforms five comparative models, demonstrating its effectiveness and accuracy in real-time apple flower detection. With its lightweight design and high accuracy, this method offers a promising solution for developing portable apple flower detection systems.

Why it matches plant phenotyping methodsリンゴ花を対象とするYOLOv8改良型の画像解析手法を開発し、検出性能を比較評価している。花の検出は開花状態や生育・収量推定に関わる植物器官の表現型取得であり、手法が研究の中心である。

abstractan improved YO-AFD method based on YOLOv8 for apple flower detection was proposed.

Retrieving the chlorophyll content of individual apple trees by reducing canopy shadow impact via a 3D radiative transfer model and UAV multispectral imagery. →

Accurate monitoring and spatial distribution of the leaf chlorophyll content (LCC) and canopy chlorophyll content (CCC) of individual apple trees are highly important for the effective management of individual plants and the promotion of the construction of modern smart orchards. However, the estimation of LCC and CCC is affected by shadows caused by canopy structure and observation geometry. In this study, we resolved the response relationship between individual apple tree crown spectra and shadows through a three-dimensional radiative transfer model (3D RTM) and unmanned aerial vehicle (UAV) multispectral images, assessed the resistance of a series of vegetation indices (VIs) to shadows and developed a hybrid inversion model that is resistant to shadow interference. The results revealed that (1) the proportion of individual tree canopy shadows exhibited a parabolic trend with time, with a minimum occurring at noon. Correspondingly, the reflectance in the visible band decreased with increasing canopy shadow ratio and reached a maximum value at noon, whereas the pattern of change in the reflectance in the near-infrared band was opposite that in the visible band. (2) The accuracy of chlorophyll content estimation varies among different VIs at different canopy shadow ratios. The top five VIs that are most resistant to changes in canopy shadow ratios are the NDVI-RE, Cire, Cigreen, TVI, and GNDVI. (3) For the constructed 3D RTM ​+ ​GPR hybrid inversion model, only four VIs, namely, NDVI-RE, Cire, Cigreen, and TVI, need to be input to achieve the best inversion accuracy. (4) Both the LCC and the CCC of individual trees had good validation accuracy (LCC: R 2 ​= ​0.775, RMSE ​= ​6.86 ​μg/cm 2 , nRMSE ​= ​12.24 ​%; CCC: R 2 ​= ​0.784, RMSE ​= ​32.33 ​μg/cm 2 , and nRMSE ​= ​14.49 ​%), and their distributions at orchard scales were characterized by considerable spatial heterogeneity. This study provides ideas for investigating the response between individual tree canopy shadows and spectra and offers a new strategy for minimizing the influence of shadow effects on the accurate estimation of chlorophyll content in individual apple trees.

Why it matches plant phenotyping methodsUAVマルチスペクトル画像と3D放射伝達モデルを用いて、個体別リンゴ樹の葉・樹冠クロロフィル含量を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。

abstractwe resolved the response relationship between individual apple tree crown spectra and shadows through a three-dimensional radiative transfer model (3D RTM) and unmanned aerial vehicle (UAV) multispectral images, assessed the resistance of a series of vegetation indices (VIs) to shadows and developed a hybrid inversion model that is resistant to shadow interference.

Exploring the impact of lenticels on the detection of soluble solids content in apples and pears using hyperspectral imaging and one-dimensional convolutional neural networks →

In this work, the effect of lenticels on the predictive performance of apple and pear soluble solids content (SSC) models developed based on hyperspectral imaging (HSI) at 380-1010 nm was investigated for the first time. Variations in the spectral properties of lenticels, pericarp, and combined lenticels and pericarp regions of interest (ROI) were analyzed using two-dimensional correlation spectroscopy method (2D-COS), factor discriminant analysis (FDA) and principal component analysis (PCA). Partial least squares regression (PLSR) was performed to develop calibration models of SSC for each ROI separately. Furthermore, variable selection algorithm and one-dimensional convolutional neural network (1D-CNN) were utilized to simplify and improve the model prediction capability. The results showed that the spectral properties of lenticels and pericarp vary considerably, while PCA could highlight the distribution of lenticels. The spectral measurement location has a significant effect on the SSC prediction accuracy. The models can be kept robust when the data sources for the prediction and calibration sets are the same. Specifically, for apple fruit, the SPA-1D-CNN achieved the best model performance with Rc² of 0.845 and Rₚ² of 0.808, respectively. For pear fruit, the best model is the CARS-1D-CNN model with Rc² of 0.887 and Rp² = 0.762. This study demonstrated that lenticels have a significant effect on model prediction performance and the 1D-CNN could be an alternative to conventional PLSR method.

Why it matches plant phenotyping methodsリンゴとナシの果実SSCという植物形質を、ハイパースペクトル画像と回帰・CNNモデルで推定し、測定領域やモデル性能の影響を検証しているため、形質取得・抽出手法が中心である。

abstractthe effect of lenticels on the predictive performance of apple and pear soluble solids content (SSC) models developed based on hyperspectral imaging (HSI) at 380-1010 nm was investigated

Integrating evolutionary algorithms and enhanced-YOLOv8 + for comprehensive apple ripeness prediction. →

The assessment of apple quality is pivotal in agricultural production management, and apple ripeness is a key determinant of apple quality. This paper proposes an approach for assessing apple ripeness from both structured and unstructured observation data, i.e., text and images. For structured text data, support vector regression (SVR) models optimized using the Whale Optimization Algorithm (WOA), Grey Wolf Optimizer (GWO), and Sparrow Search Algorithm (SSA) were utilized to predict apple ripeness, with the WOA-optimized SVR demonstrating exceptional generalization capabilities. For unstructured image data, an Enhanced-YOLOv8+, a modified YOLOv8 architecture integrating Detect Efficient Head (DEH) and Efficient Channel Attention (ECA) mechanism, was employed for precise apple localization and ripeness identification. The synergistic application of these methods resulted in a significant improvement in prediction accuracy. These approaches provide a robust framework for apple quality assessment and deepen the understanding of the relationship between apple maturity and observed indicators, facilitating more informed decision-making in postharvest management.

Why it matches plant phenotyping methods画像と最適化アルゴリズムを用いてリンゴ果実の成熟度を推定する手法が中心であり、果実の状態を直接評価する植物フェノタイピングとして適格。

abstractThis paper proposes an approach for assessing apple ripeness from both structured and unstructured observation data, i.e., text and images.

Comprehensive Performance Evaluation of YOLOv12, YOLO11, YOLOv10, YOLOv9 and YOLOv8 on Detecting and Counting Fruitlet in Complex Orchard Environments →

This study systematically performed an extensive real-world evaluation of the performances of all configurations of YOLOv8, YOLOv9, YOLOv10, YOLO11( or YOLOv11), and YOLOv12 object detection algorithms in terms of precision, recall, mean Average Precision at 50\% Intersection over Union (mAP@50), and computational speeds including pre-processing, inference, and post-processing times immature green apple (or fruitlet) detection in commercial orchards. Additionally, this research performed and validated in-field counting of the fruitlets using an iPhone and machine vision sensors. Among the configurations, YOLOv12l recorded the highest recall rate at 0.90, compared to all other configurations of YOLO models. Likewise, YOLOv10x achieved the highest precision score of 0.908, while YOLOv9 Gelan-c attained a precision of 0.903. Analysis of mAP@0.50 revealed that YOLOv9 Gelan-base and YOLOv9 Gelan-e reached peak scores of 0.935, with YOLO11s and YOLOv12l following closely at 0.933 and 0.931, respectively. For counting validation using images captured with an iPhone 14 Pro, the YOLO11n configuration demonstrated outstanding accuracy, recording RMSE values of 4.51 for Honeycrisp, 4.59 for Cosmic Crisp, 4.83 for Scilate, and 4.96 for Scifresh; corresponding MAE values were 4.07, 3.98, 7.73, and 3.85. Similar performance trends were observed with RGB-D sensor data. Moreover, sensor-specific training on Intel Realsense data significantly enhanced model performance. YOLOv11n achieved highest inference speed of 2.4 ms, outperforming YOLOv8n (4.1 ms), YOLOv9 Gelan-s (11.5 ms), YOLOv10n (5.5 ms), and YOLOv12n (4.6 ms), underscoring its suitability for real-time object detection applications.

Why it matches plant phenotyping methods果実数という植物器官形質の画像ベース推定を対象に、複数YOLOモデルの性能比較とiPhone・RGB-Dセンサーによる圃場カウント検証を中心的に行っているため。

abstractThis study systematically performed an extensive real-world evaluation of the performances of all configurations of YOLOv8, YOLOv9, YOLOv10, YOLO11( or YOLOv11), and YOLOv12 object detection algorithms

CNN-BASED DETECTION OF POWDERY MILDEW AND RUST IN APPLE ORCHARDS FOR OPTIMIZING CROP MANAGEMENT →

In many parts of India, apple trees are among the most popular crops. Large amounts of apples are exported annually, which has a major positive impact on the country's economy. However, a number of diseases in apple trees are common. They indicate a significant risk to apple production and lead to significant financial losses for producers. These diseases mostly affect the leaves of apple plants. In a country where a significant portion of the workforce is employed in agriculture, prompt identification and management of such diseases are essential. It used to take a lot of time and effort to diagnose diseases in apple plants via laboratory testing. Machine Learning (ML) methods offer a fast and accurate detection of diseased leaves in the apple orchard. This study aimed to develop a robust Convolutional Neural Network (CNN) model for identifying apple leaf diseases. A dataset comprising 1,532 images categorized into Healthy, Powdery mildew, and Rust classes was used. The CNN model consisted of six convolutional layers, six max-pooling layers, a flatten layer, and fully connected layers. Images were pre-processed (resized to 256x256 pixels, normalized, and augmented) to improve computational efficiency. The model was evaluated using metrics such as accuracy, precision, recall, F1-score, and a confusion matrix. The model achieved a training accuracy of 98.02%, validation accuracy of 85.17%, and overall accuracy of 91.34%. Precision and recall for individual classes ranged from 86.05% to 96.55%. F1-scores showed balanced performance across categories, with a weighted average of 92.54%. These results demonstrate the model's effectiveness in classifying leaf conditions and its potential to enhance disease management in apple orchards and similar crops. Keywords: Machine Learning, Apple Orchard, Crop Management, Convolutional Neural Network, Evaluation Metrics

Why it matches plant phenotyping methodsリンゴ葉の病害状態を画像から分類するCNN手法の開発と評価が研究の中心であり、植物病害フェノタイプの取得・推定に該当する。

abstractThis study aimed to develop a robust Convolutional Neural Network (CNN) model for identifying apple leaf diseases.

Plant leaf disease detection using local binary pattern and deep convolutional neural networks →

Plants are susceptible to pathogen infections during their growing period leading to reduced crop quality and yield. Traditional disease detection methods such as expert diagnosis and pathogen analysis rely on experienced professionals and could be time-consuming and prone to errors. Deep convolutional neural networks (CNNs) have exhibited their potential to detect plant diseases on the basis of visual patterns of leaves. Most of the existing CNN based methods do not take advantage of additional information. Most of the disease significantly affects the texture of the plant leaves. Therefore, texture features can provide complementary information to get better results. In this paper, local binary pat tern (LBP) technique is used to extract texture information that is stacked with original image. A CNN model is proposed that takes embedded texture and spectral information to detect crop diseases using leaf images. The experiments are carried out on Apple, Corn, and Potato crops from Plant Village dataset. The proposed method achieved the overall accuracy up to 98.73% (κ = 98.04). It is found that LBP makes significant difference in disease classification accuracy and helps the proposed method exhibit better performance than some existing well known CNN models.

Why it matches plant phenotyping methods葉画像から植物病害状態を推定するLBP＋CNN手法を開発・評価しており、植物表現型（病害）の取得・抽出が中心である。

abstractIn this paper, local binary pat tern (LBP) technique is used to extract texture information that is stacked with original image.

Detection of Apple Leaf Diseases Based on LightYOLO-AppleLeafDx. →

Early detection of apple leaf diseases is essential for enhancing orchard management efficiency and crop yield. This study introduces LightYOLO-AppleLeafDx, a lightweight detection framework based on an improved YOLOv8 model. Key enhancements include the incorporation of Slim-Neck, SPD-Conv, and SAHead modules, which optimize the model's structure to improve detection accuracy and recall while significantly reducing the number of parameters and computational complexity. Ablation studies validate the positive impact of these modules on model performance. The final LightYOLO-AppleLeafDx achieves a precision of 0.930, mAP@0.5 of 0.965, and mAP@0.5:0.95 of 0.587, surpassing the original YOLOv8n and other benchmark models. The model is highly lightweight, with a size of only 5.2 MB, and supports real-time detection at 107.2 frames per second. When deployed on an RV1103 hardware platform via an NPU-compatible framework, it maintains a detection speed of 14.8 frames per second, demonstrating practical applicability. These results highlight the potential of LightYOLO-AppleLeafDx as an efficient and lightweight solution for precision agriculture, addressing the need for accurate and real-time apple leaf disease detection.

Why it matches plant phenotyping methodsリンゴ葉の病害状態を画像から検出する軽量YOLOフレームワークの開発とアブレーション・ベンチマーク検証が中心であり、植物病害表現型の取得手法に該当する。

abstractThis study introduces LightYOLO-AppleLeafDx, a lightweight detection framework based on an improved YOLOv8 model.

Plant Disease Detection Using Hybrid MobileNetV2- Compact CNN Architecture with LIME Integration →

This paper presents an advanced approach to plant disease detection by implementing explainable AI techniques that combine MobileNetV2 architecture with transfer learning and compact convolutional neural networks (CNN). The study compares three distinct models' performance on a plant leaf disease dataset, revealing MobileNetV2's superior accuracy of 95% with 94% precision in disease classification, despite requiring 850 seconds for training. The Compact CNN achieved 82% accuracy with minimal training time of 420 seconds, demonstrating its efficiency for resource-constrained applications. Disease-specific analysis showed exceptional detection rates for common plant diseases, with Apple Scab at 96.5%, Black Rot at 94.8%, and Cedar Rust at 95.2%. The integration of LIME (Local Interpretable Model-agnostic Explanations) provided transparent insights into the model's decision-making process, while the Compact CNN demonstrated 45% reduced memory usage compared to MobileNetV2. This implementation establishes a robust framework for practical agricultural applications, balancing high accuracy with computational efficiency and interpretability.

Why it matches plant phenotyping methods植物葉画像から病害状態を推定するCNN手法を比較・評価し、LIMEによる解釈性と計算効率も検討しており、病害フェノタイピング手法が中心である。

abstractThis paper presents an advanced approach to plant disease detection by implementing explainable AI techniques that combine MobileNetV2 architecture with transfer learning and compact convolutional neural networks (CNN).

Exploring the impact of lenticels on the detection of soluble solids content in apples and pears using hyperspectral imaging and one-dimensional convolutional neural networks. →

In this work, the effect of lenticels on the predictive performance of apple and pear soluble solids content (SSC) models developed based on hyperspectral imaging (HSI) at 380-1010 nm was investigated for the first time. Variations in the spectral properties of lenticels, pericarp, and combined lenticels and pericarp regions of interest (ROI) were analyzed using two-dimensional correlation spectroscopy method (2D-COS), factor discriminant analysis (FDA) and principal component analysis (PCA). Partial least squares regression (PLSR) was performed to develop calibration models of SSC for each ROI separately. Furthermore, variable selection algorithm and one-dimensional convolutional neural network (1D-CNN) were utilized to simplify and improve the model prediction capability. The results showed that the spectral properties of lenticels and pericarp vary considerably, while PCA could highlight the distribution of lenticels. The spectral measurement location has a significant effect on the SSC prediction accuracy. The models can be kept robust when the data sources for the prediction and calibration sets are the same. Specifically, for apple fruit, the SPA-1D-CNN achieved the best model performance with R c 2 of 0.845 and R p 2 of 0.808, respectively. For pear fruit, the best model is the CARS-1D-CNN model with Rc 2 of 0.887 and Rp 2 = 0.762. This study demonstrated that lenticels have a significant effect on model prediction performance and the 1D-CNN could be an alternative to conventional PLSR method.

Why it matches plant phenotyping methodsハイパースペクトル画像からリンゴ・ナシ果実の可溶性固形分含量を推定するモデルを開発・比較しており、形質取得・推定手法が研究の中心である。

abstractapple and pear soluble solids content (SSC) models developed based on hyperspectral imaging (HSI) at 380-1010 nm was investigated

A refined DenseNet Deep Learning Network for Apple Leaf Disease Prediction →

Revolutionary strategies for managing agricultural diseases have been made possible by recent developments in machine learning, especially in apple leaf disease prediction. According to recent research, convolutional neural networks with fewer connections between layers near their inputs and those near the output can be trained with far greater depth, accuracy, and efficiency. To make a more precise diagnosis of apple-leaf defects than existing architectures, this work proposed a method combining DenseNet-121 and optimising transfer learning strategy for multiclass classification. DenseNet-121 is used as a feature extractor as it strengthens feature propagation and reuse, leading to sustainable feature parameter reduction. The experiment is performed on 3 publicly accessible datasets with 3 classes, 6 classes and 9 classes of apple disease in leaf. The network architecture is fed with augmented data to avoid the problem of class imbalance. The proposed model has responded exceptionally well on all three datasets, claiming 99.9%, 99% and 96% accuracy. Comparative studies and experimental data demonstrate the competitive prediction accuracy of the suggested approach.

Why it matches plant phenotyping methodsリンゴ葉の病徴を画像から分類・予測する深層学習手法の開発と比較評価が中心であり、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採用する。

titleA refined DenseNet Deep Learning Network for Apple Leaf Disease Prediction

Apple tree architectural trait phenotyping with organ-level instance segmentation from point cloud →

Three-dimensional (3D) plant phenotyping techniques measure organ-level traits effectively and provide detailed plant growth information to breeders. In apple tree breeding, architectural traits can determine photosynthesis efficiency and characterize the developmental stages of trees. The overall goal of this study was to develop a deep learning-based organ-level instance segmentation method to quantify the 3D architectural traits of apple trees. This study utilized PointNeXt for the semantic segmentation of apple tree point clouds, classifying them into trunks and branches, and benchmarked its performance against several competitive models, including PointNet, PointNet++, and Point Transformer V2 (PTv2). A cylinder-based constraint method was introduced to refine the semantic segmentation results. Next, the branches were identified with the density-based spatial clustering of applications with noise (DBSCAN) algorithm. The type of 3D skeleton vertices determined whether a cluster represented a single branch or multiple branches. If multiple, a graph-based technique further separated them. This study also directly applied the instance segmentation model SoftGroup++ to the apple tree point clouds and analyzed the segmentation results on the apple tree dataset. Finally, seven architectural traits of apple trees were extracted, including height, volume, and crown width of the tree, as well as height and diameter for the trunk, and length and count for the branches. The experimental results showed that the post-processed mIoU values for PointNet, PointNet++, PTv2, and PointNeXt were 0.8495, 0.8535, 0.9500, and 0.9481, respectively. The final instance segmentation results based on SoftGroup++ and PointNeXt achieved mAP_50 of 0.815 and 0.842, respectively. For traits such as tree height, trunk length and diameter, branch length, and branch count, the method based on PointNeXt achieved R² values of 0.987, 0.788, 0.877, 0.796, and 0.934, with mean absolute percentage errors of 0.86 %, 2.17 %, 5.93 %, 10.24 %, and 13.55 %, respectively. The segmentation results of PTv2 and SoftGroup++ were also used to extract the phenotypic traits of apple trees, achieving results comparable to those of PointNeXt. The proposed method demonstrates a cost-effective and accurate approach for extracting the architectural traits of apple trees, which will benefit apple breeding programs as well as the precision management of apple orchards.

Why it matches plant phenotyping methodsリンゴ樹の点群から器官レベルのインスタンスセグメンテーションにより3D建築形質を抽出する手法を開発・比較検証しており、植物フェノタイピング手法が研究の中心である。

abstractThe overall goal of this study was to develop a deep learning-based organ-level instance segmentation method to quantify the 3D architectural traits of apple trees.

Estimating optimal crop-load for individual branches in apple tree canopies using YOLOv8 →

Shortage of labor in fruit crop production has become a significant challenge in recent years. Therefore, mechanized and automated machines have emerged as promising alternatives to labor-intensive orchard operations such as harvesting, pruning, and thinning. One of the key aspects of the automated machines in accomplishing these tasks is their ability to identify tree canopy parts such as trunk and branches and estimate their geometric and topological parameters such as branch diameter, branch length, branch angles, and spacing between branches. By utilizing geometric parameters such as branch diameter, length, and orientation, researchers then can develop automated pruning and thinning systems that make more effective decisions to achieve optimal fruit yield and quality by accurately estimating the desired crop-load. In this study, we propose a machine vision system for estimating one of the canopy parameters in apple orchards: branch diameter. This parameter was used to estimate the optimal number of fruit that individual branches could bear in a commercial orchard, which provides a basis for robotic pruning, flower thinning, and fruitlet thinning so that desired fruit yield and quality could be achieved. Utilizing color and depth information collected with an RGB-D sensor (Azure Kinect DK, Microsoft, Redmond, WA), a YOLOv8-based instance segmentation technique was developed to identify trunks and branches of apple trees in the dormant season. We then applied a Principal Component Analysis (PCA) technique to estimate branch orientation, which was subsequently utilized to estimate branch diameter. The estimated branch diameter was used to calculate the Limb Cross Sectional Area (LCSA), which was then used to estimate optimal crop-load, as a larger LCSA indicates a higher potential fruit-bearing capacity of the branch. With this approach, Root Mean Squared Error (RMSE) for branch diameter estimation was calculated to be 2.06 mm (relative RMSE 10.82%) and the same for crop-load estimation (Number of fruits per branch) to be 3.93 (relative RMSE 22.25%). Our study demonstrated a promising workflow with a high level of performance in identifying and sizing branches of apple trees in a dynamic orchard environment and integrating farm management practices into automated decision-making for optimizing crop-load in apple orchards.

Why it matches plant phenotyping methodsRGB-D画像とYOLOv8・PCAを用いてリンゴ枝径および枝ごとの作物負荷を推定する手法を開発・評価しており、植物形質の取得が中心です。

abstractIn this study, we propose a machine vision system for estimating one of the canopy parameters in apple orchards: branch diameter.

Apple tree architectural trait phenotyping with organ-level instance segmentation from point cloud →

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

Why it matches plant phenotyping methodsリンゴ樹の器官レベルインスタンスセグメンテーションにより樹体構造形質を抽出する手法であり、植物フェノタイピング手法が中心と明示されている。

titleApple tree architectural trait phenotyping with organ-level instance segmentation from point cloud

A systematic review of deep learning techniques for apple leaf diseases classification and detection. →

Agriculture sustains populations and provides livelihoods, contributing to socioeconomic growth. Apples are one of the most popular fruits and contains various antioxidants that reduce the risk of chronic diseases. Additionally, they are low in calories, making them a healthy snack option for all ages. However, several factors can adversely affect apple production. These issues include diseases that drastically lower yield and quality and cause farmers to lose millions of dollars. To minimize yield loss and economic effects, it is essential to diagnose apple leaf diseases accurately and promptly. This allows targeted pesticide and insecticide use. However, farmers find it difficult to distinguish between different apple leaf diseases since their symptoms are quite similar. Computer vision applications have become an effective tool in recent years for handling these issues. They can provide accurate disease detection and classification through massive image datasets. This research analyzes and evaluates datasets, deep learning methods and frameworks built for apple leaf disease detection and classification. A systematic analysis of 45 articles published between 2016 and 2024 was conducted to evaluate the latest developments, approaches, and research needs in this area.

Why it matches plant phenotyping methodsリンゴ葉の病徴を画像から検出・分類する深層学習手法を体系的に評価したレビューであり、植物の病害状態を観測するフェノタイピング手法が中心です。

titleA systematic review of deep learning techniques for apple leaf diseases classification and detection.

Fruit Detection and Yield Mass Estimation from a UAV Based RGB Dense Cloud for an Apple Orchard →

Precise photogrammetric mapping of preharvest conditions in an apple orchard can help determine the exact position and volume of single apple fruits. This can help estimate upcoming yields and prevent losses through spatially precise cultivation measures. These parameters also are the basis for effective storage management decisions, post-harvest. These spatial orchard characteristics can be determined by low-cost drone technology with a consumer grade red-green-blue (RGB) sensor. Flights were conducted in a specified setting to enhance the signal-to-noise ratio of the orchard imagery. Two different altitudes of 7.5 m and 10 m were tested to estimate the optimum performance. A multi-seasonal field campaign was conducted on an apple orchard in Brandenburg, Germany. The test site consisted of an area of 0.5 ha with 1334 trees, including the varieties ‘Gala’ and ‘Jonaprince’. Four rows of trees were tested each season, consisting of 14 blocks with eight trees each. Ripe apples were detected by their color and structure from a photogrammetrically created three-dimensional point cloud with an automatic algorithm. The detection included the position, number, volume and mass of apples for all blocks over the orchard. Results show that the identification of ripe apple fruit is possible in RGB point clouds. Model coefficients of determination ranged from 0.41 for data captured at an altitude of 7.5 m for 2018 to 0.40 and 0.53 for data from a 10 m altitude, for 2018 and 2020, respectively. Model performance was weaker for the last captured tree rows because data coverage was lower. The model underestimated the number of apples per block, which is reasonable, as leaves cover some of the fruits. However, a good relationship to the yield mass per block was found when the estimated apple volume per block was combined with a mean apple density per variety. Overall, coefficients of determination of 0.56 (for the 7.5 m altitude flight) and 0.76 (for the 10 m flights) were achieved. Therefore, we conclude that mapping at an altitude of 10 m performs better than 7.5 m, in the context of low-altitude UAV flights for the estimation of ripe apple parameters directly from 3D RGB dense point clouds.

Why it matches plant phenotyping methodsUAV RGB三次元点群から果実の位置・数・体積・質量を自動推定し、飛行高度別の性能を検証しており、植物形質の取得・推定手法が中心である。

abstractRipe apples were detected by their color and structure from a photogrammetrically created three-dimensional point cloud with an automatic algorithm.

Fruit Detection and Yield Mass Estimation from a UAV Based RGB Dense Cloud for an Apple Orchard →

Precise photogrammetric mapping of preharvest conditions in an apple orchard can help determine the exact position and volume of single apple fruits. This can help estimate upcoming yields and prevent losses through spatially precise cultivation measures. These parameters also are the basis for effective storage management decisions, post-harvest. These spatial orchard characteristics can be determined by low-cost drone technology with a consumer grade red-green-blue (RGB) sensor. Flights were conducted in a specified setting to enhance the signal-to-noise ratio of the orchard imagery. Two different altitudes of 7.5 m and 10 m were tested to estimate the optimum performance. A multi-seasonal field campaign was conducted on an apple orchard in Brandenburg, Germany. The test site consisted of an area of 0.5 ha with 1334 trees, including the varieties ‘Gala’ and ‘Jonaprince’. Four rows of trees were tested each season, consisting of 14 blocks with eight trees each. Ripe apples were detected by their color and structure from a photogrammetrically created three-dimensional point cloud with an automatic algorithm. The detection included the position, number, volume and mass of apples for all blocks over the orchard. Results show that the identification of ripe apple fruit is possible in RGB point clouds. Model coefficients of determination ranged from 0.41 for data captured at an altitude of 7.5 m for 2018 to 0.40 and 0.53 for data from a 10 m altitude, for 2018 and 2020, respectively. Model performance was weaker for the last captured tree rows because data coverage was lower. The model underestimated the number of apples per block, which is reasonable, as leaves cover some of the fruits. However, a good relationship to the yield mass per block was found when the estimated apple volume per block was combined with a mean apple density per variety. Overall, coefficients of determination of 0.56 (for the 7.5 m altitude flight) and 0.76 (for the 10 m flights) were achieved. Therefore, we conclude that mapping at an altitude of 10 m performs better than 7.5 m, in the context of low-altitude UAV flights for the estimation of ripe apple parameters directly from 3D RGB dense point clouds.

Why it matches plant phenotyping methodsUAV RGB三次元点群からリンゴ果実の位置・個数・体積・質量を自動推定し、飛行高度別に性能検証しており、植物形質取得手法が中心である。

abstractRipe apples were detected by their color and structure from a photogrammetrically created three-dimensional point cloud with an automatic algorithm.

A hybrid Framework for plant leaf disease detection and classification using convolutional neural networks and vision transformer →

Recently, scientists have widely utilized Artificial Intelligence (AI) approaches in intelligent agriculture to increase the productivity of the agriculture sector and overcome a wide range of problems. Detection and classification of plant diseases is a challenging problem due to the vast numbers of plants worldwide and the numerous diseases that negatively affect the production of different crops. Early detection and accurate classification of plant diseases is the goal of any AI-based system. This paper proposes a hybrid framework to improve classification accuracy for plant leaf diseases significantly. This proposed model leverages the strength of Convolutional Neural Networks (CNNs) and Vision Transformers (ViT), where an ensemble model, which consists of the well-known CNN architectures VGG16, Inception-V3, and DenseNet20, is used to extract robust global features. Then, a ViT model is used to extract local features to detect plant diseases precisely. The performance proposed model is evaluated using two publicly available datasets (Apple and Corn). Each dataset consists of four classes. The proposed hybrid model successfully detects and classifies multi-class plant leaf diseases and outperforms similar recently published methods, where the proposed hybrid model achieved an accuracy rate of 99.24% and 98% for the apple and corn datasets.

Why it matches plant phenotyping methods植物葉の病害状態を画像から検出・分類するCNN/ViT手法の開発と評価が研究の中心であり、植物表現型の取得・推定に該当する。

abstractThis paper proposes a hybrid framework to improve classification accuracy for plant leaf diseases significantly.

SVM-RBN Model with Attentive Feature Culling Method for Early Detection of Fruit Plant Diseases →

Accurate disease identification and early disease management strategies are required in India to achieve high production standards and good quality in fruits and vegetables. Image-based evaluations approaches have evolved nowadays as a result of technical developments. However, producing the wrong decision may have a negative impact on productivity. Thus, this study offered hybrid SVM-RBN model with attentive feature culling method for automatically recognizing diseases in apple fruits with high accuracy. As a result, the model generated more effective outcomes with 96% accuracy, 99% precision, 94% recall, and 93% F1 Score. Thus, by employing this technology, one may detect fruit plant illnesses at an early stage, thereby increasing fruit yield.

Why it matches plant phenotyping methodsリンゴ果実の画像から病害を自動認識するモデルを開発・評価しており、植物の病害状態を抽出する画像ベースのフェノタイピング手法が中心です。

abstractthis study offered hybrid SVM-RBN model with attentive feature culling method for automatically recognizing diseases in apple fruits with high accuracy

Differential scanning calorimetry: a novel approach to study spring frost tolerance in pome tree floral tissues →

Spring frosts occur in many pome fruit-producing regions globally. They are highly detrimental to floral tissues and yield, making frost tolerance of the reproductive organs one of the major breeding challenges. Currently, frost tolerance of flowers and floral buds is determined by meticulous observations of frost damage symptoms, or by methods that relate the damage with the accumulation of physiologically-relevant biochemicals. These methods are often inaccurate and are only feasible in instances of severe frost. We propose differential scanning calorimetry (DSC) to assess the frost tolerance of pome fruit floral tissues by measuring the heat flow of tissue samples when passing the freezing transition. DSC was applied to floral organs isolated from recently open king flowers of the apple 'Jonagold' (Malus domestica Borkh.) and the European pear 'Conference' (Pyrus communis L.) to simulate frost and determine freezing temperature as a quantitative indicator of frost tolerance. Freezing, crystallization, and melting points were measured by cooling and heating isolated ovules, stamens, and stigmas of mature king flowers from 20 to −40 °C and back to 20 °C. In pear, tissue-specific effects were observed, with ovules showing the highest (−9.6 °C) and stamens showing the lowest average freezing temperature (−13.1 °C), indicating that stamens are less susceptible to frost. In contrast, no significant differences in frost tolerance were detected among apple organs (freezing temperature of −11.1 °C). This study shows that DSC is an efficient method for monitoring and evaluating frost tolerance in pome fruit floral tissues and could be utilized for high-throughput phenotyping in breeding programs.

Why it matches plant phenotyping methodsDSCによる花器官の凍結温度測定を、霜害耐性の定量的な表現型取得法として提案・評価しており、方法開発と検証が中心である。

abstractWe propose differential scanning calorimetry (DSC) to assess the frost tolerance of pome fruit floral tissues by measuring the heat flow of tissue samples when passing the freezing transition.

Panoptic segmentation for complete labeling of fruit microstructure in 3D micro-CT images with deep learning →

Metabolic processes in plant organs involving transport of water, metabolic gasses, and nutrients depend on the three-dimensional (3D) microscopic tissue morphology. However, imaging and quantifying this microstructure, including the spatial layout of parenchyma cells, pores, vascular bundles and special features such as stone cell clusters (brachysclereids), is challenging. To address this, a 3D deep learning-based panoptic segmentation model, combining semantic and instance segmentation, was developed to accelerate and improve microstructure characterization of apple and pear fruit tissue in X-ray micro-computed tomography (CT) images. In addition, various training datasets and data augmentation techniques, including synthetic data, were explored to enhance segmentation quality. The 3D panoptic segmentation achieved an Aggregated Jaccard Index of 0.89 and 0.77 for apple and pear tissue, respectively, outperforming both the previously designed 2D instance segmentation model and a marker-based watershed segmentation benchmark. The model successfully labelled vascular bundles with a Dice Similarity Coefficient (DSC) of 0.51 in apple tissue and 0.79 in pear tissue, although thin vasculature in apple remained more challenging to segment. The 3D panoptic segmentation model achieved a DSC of 0.81 and effectively segmented stone cell clusters in pear tissue. Despite evaluating different methods to enhance segmentation quality, none improved test performance beyond that of the model trained on the standard dataset. The proposed 3D panoptic segmentation model offers the most complete automated protocol to date for plant tissue labelling and morphometric quantification from native X-ray micro-CT images, without extensive sample preparation such as contrast labelling. The developed method, if not replaces, drastically accelerates conventional human-in-the-loop analysis of such images.

Why it matches plant phenotyping methods植物果実組織の3D微細構造をマイクロCT画像から自動抽出・定量化する深層学習手法を開発し、既存手法およびベンチマークと性能比較しているため、植物フェノタイピング手法が研究の中心です。

abstracta 3D deep learning-based panoptic segmentation model, combining semantic and instance segmentation, was developed to accelerate and improve microstructure characterization of apple and pear fruit tissue in X-ray micro-computed tomography (CT) images.

Multispectral imaging-based detection of apple bruises using segmentation network and classification model. →

Bruises can affect the appearance and nutritional value of apples and cause economic losses. Therefore, the accurate detection of bruise levels and bruise time of apples is crucial. In this paper, we proposed a method that combines a self-designed multispectral imaging system with deep learning to accurately detect the level and time of bruising on apples. To enhance the accuracy of extracting bruised regions with subtle features and irregular edges, an improved DeepLabV3+ was proposed. More specifically, depthwise separable convolution and efficient channel attention were employed, and the loss function was replaced with a focal loss. With these improvements, DeepLabV3+ achieved the maximum intersection over union of 95.5% and 91.0% for segmenting bruises on two types of apples in the test set, as well as maximum F1-score of 97.5% and 95.2%. In addition, the spectral data of the bruised regions were extracted. After spectral preprocessing, EfficientNetV2, DenseNet121, and ShuffleNetV2 were utilized to identify the bruise levels and times and DenseNet121 exhibited the best performance. To improve the identification accuracy, an improved DenseNet121 was proposed. The learning rate was adjusted using the cosine annealing algorithm, and squeeze-and-excitation attention mechanism and the Gaussian error linear unit activation function were utilized. Test set results demonstrated that the accuracies of the bruising levels were 99.5% and 99.1%, and those of the bruise time were 99.0% and 99.3%, respectively. This provides a new method for detecting bruise levels and bruised time on apples.

Why it matches plant phenotyping methodsリンゴの打撲領域・程度・経過時間を、多波長画像、セグメンテーション、スペクトル解析、深層学習で推定する方法を開発・評価しており、植物器官の状態計測が研究の中心である。

abstractwe proposed a method that combines a self-designed multispectral imaging system with deep learning to accurately detect the level and time of bruising on apples.

ET-PatchNet: A low-memory, efficient model for Multi-view Stereo with a case study on the 3D reconstruction of fruit tree branches →

The achievement of robotic fruit harvesting in intelligent farming depends heavily on the precise reconstruction of tree branch structures that direct harvesting arm movements. However, existing research in this field often grapples with computational inefficiencies and high costs. In response, we designed ET-PatchNet, a low-memory neural network for generating depth maps within the Multi-view Stereo (MVS) process, enabling efficient 3D reconstruction of branches. This highly Efficient network is based on Transformer and Patchmatchnet. ET-PatchNet incorporates an efficient backbone that includes self-attention and cross-attention mechanisms based on Transformer, which enriches global and 3D consistency information and enhances depth prediction accuracy and generalization performance. Furthermore, an adaptive depth resampling method has been developed, which is embedded in a iterative, coarse-to-fine, depth regression architecture based on learnable patches to minimize memory usage. To further amplify the representation capacity of depth characteristics, an auxiliary task has been integrated. Experimental results show that ET-PatchNet outperforms its competitors in completeness, computational efficiency, and low memory usage in evaluating the DTU and Tanks&Temples datasets. When predicting a single depth map at a resolution of 1152 × 864 pixels, it only took 0.13 s to inference, with a memory usage of just 2824MB. Moreover, the 3D structure of observable branches on apple trees has been effectively reconstructed by fine-tuning our model on the BlendedMVS dataset. The mean and variance of distances between our reconstructed branch points and reference points are only 0.0292mm and 0.0187mm2. In conclusion, ET-PatchNet is ideal for integration into mobile embedded fruit harvesting equipment and exhibits significant potential for a wide range of applications.

Why it matches plant phenotyping methods果樹枝の3D構造を抽出するMVS手法を開発し、リンゴ樹で精度評価しており、植物形態計測が中心的な技術貢献である。

abstractwe designed ET-PatchNet, a low-memory neural network for generating depth maps within the Multi-view Stereo (MVS) process, enabling efficient 3D reconstruction of branches.

Three-dimensional Quantification of Apple Phenotypic Traits based on Deep Learning Instance Segmentation →

Estimating three-dimensional (3D) fruit-level phenotypic traits of apple trees can potentially improve apple orchard breeders' management strategy. However, the phenotypic traits of apple fruits (including the quantity, 3D distribution, and volume of apples) constitute important parameters that influence yield but are difficult to quantify manually. Therefore, it is necessary to effectively and efficiently quantify apple phenotyping to monitor apple yield and support a better management system. This study developed a novel method for extracting individual 3D apple traits and 3D mapping for three apple training systems. The 3D point cloud of apples was reconstructed from multi-view images collected via a multi-camera system-based unmanned-aerial vehicle. Individual apples in the 3D point cloud were extracted via a 3D instance segmentation algorithm that included generalized sparse convolutional neural networks, a weighted discriminative loss function, and a varying density-based 3D clustering method. The developed apple trait extraction algorithm can help compute the position and volume of an individual apple. The R² values with the average weighted-mean-absolute-percentage error (WMAPE) of apple counting and apple volume estimation were 0.84 - 0.99 (VMAPE = 3.41 - 12.75%) and 0.84 - 0.90 (WMAPE = 4.74 - 7.03%), respectively. The 3D spatial and volumetric distribution of apples were obtained and analyzed. This study developed an effective method combining 3D photography and 3D instance segmentation that can accurately estimate individual apple phenotypic traits from different types of apple training systems in orchards and can also be utilized for the analysis of other fruit traits.

Why it matches plant phenotyping methodsリンゴ果実の個体数・3D分布・体積という植物形質を、マルチビュー画像、3D再構成、インスタンスセグメンテーションで抽出する手法を開発・検証しており、フェノタイピング手法が研究の中心である。

abstractThis study developed a novel method for extracting individual 3D apple traits and 3D mapping for three apple training systems.

Monitoring apple flowering date at 10 m spatial resolution based on crop reference curves →

Apple cultivation is a mainstay industry that promotes agricultural development and boosts farmers' income in Shaanxi Province. Monitoring the flowering date of apple trees is essential for frost damage prevention and yield assessment. However, conventional ground survey methods suffer from high costs and low accuracy, and traditional approaches relying on meteorological data have limitations in spatial resolution. In this study, a set of Normalized Difference Vegetation Index (NDVI) time series, referred to as Crop Reference Curves (CRC), was extracted from pure apple tree MODIS pixels. Subsequently, this CRC was utilized to reconstruct daily 10 m NDVI data from Sentinel-2 imagery. By comparing the spatial phenological variances between the CRC and the reconstructed NDVI sequence, the historical apple flowering date was monitored and mapped with a 10 m spatial resolution in Shaanxi Province. Furthermore, we compared and analyzed the effects of Sentinel-2 images input number (5, 6, 8, and 9 scenes) on the accuracy of flowering monitoring. The results revealed that the scheme using 8 images with an average annual distribution yielded an absolute error of 2 days in monitoring the flowering date in six counties of Yan'an in 2019, indicating an effective fitting effect on monitoring the apple flowering date. The scheme employing 9 images achieved an absolute error of 1.33 days, offering the highest precision in monitoring apple flowering date. Furthermore, when using 9 images, the average error in flowering monitoring remained within 2 days from 2019 to 2021 in four validation study areas, demonstrating strong fitting and practical applicability for monitoring apple flowering dates. This method can be utilized for rapid, efficient and high-precision monitoring of apple flowering date in a wide range with a 10 m spatial resolution. Additionally, the analysis of reconstructed NDVI characteristic can serve as a technical reference for fruit forest classification and growth trend prediction.

Why it matches plant phenotyping methodsリンゴ樹の開花日という植物フェノタイプを、NDVI時系列とSentinel-2画像から推定・地図化する手法を開発し、画像枚数と精度を比較検証しており、フェノタイピング手法が中心である。

abstractIn this study, a set of Normalized Difference Vegetation Index (NDVI) time series, referred to as Crop Reference Curves (CRC), was extracted from pure apple tree MODIS pixels. Subsequently, this CRC was utilized to reconstruct daily 10 m NDVI data from Sentinel-2 imagery.

Crop Disease Detection by Deep Joint Segmentation and Hybrid Classification Model: A CAD‐Based Agriculture Development System →

Precise detection of crop disease at the early stage is a crucial task, which will reduce the spreading of disease by taking preventive measures. The main goal of this research is to propose a hybrid classification system for detecting crop disease by utilising Modified Deep Joint (MDJ) segmentation. The detection of crop diseases involves five stages. They are data acquisition, pre‐processing, segmentation, feature extraction and disease detection. In the initial stage, image data of diverse crops is gathered in the data acquisition phase. According to the work, we are considering Apple and corn crops with benchmark datasets. The input image is subjected to pre‐processing by utilising the median filtering process. Subsequently, the pre‐processed image under goes a segmentation process, where Modified Deep Joint segmentation is proposed in this work. From the segmented image, features like shape, colour, texture‐based features and Improved Median Binary Pattern (IMBP)‐based features are extracted. Finally, the extracted features are given to the hybrid classification system for identifying the crop diseases. The hybrid classification model includes Bidirectional Long Short‐Term Memory (Bi‐LSTM) and Deep Belief Network (DBN) classifiers. The outcome of both the classifiers is the score, which is subjected to an improved score level fusion model, which determines the final detection results. Finally, the performance of the proposed hybrid model is evaluated over existing methods for various metrics. At a training data of 90%, the proposed scheme attained an accuracy of 0.965, while conventional methods achieved less accuracy rates.

Why it matches plant phenotyping methods画像から作物病害の状態を推定するセグメンテーション・特徴抽出・分類手法の開発と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 main goal of this research is to propose a hybrid classification system for detecting crop disease by utilising Modified Deep Joint (MDJ) segmentation.

Geometric Feature Characterization of Apple Trees from 3D LiDAR Point Cloud Data. →

The geometric feature characterization of fruit trees plays a role in effective management in orchards. LiDAR (light detection and ranging) technology for object detection enables the rapid and precise evaluation of geometric features. This study aimed to quantify the height, canopy volume, tree spacing, and row spacing in an apple orchard using a three-dimensional (3D) LiDAR sensor. A LiDAR sensor was used to collect 3D point cloud data from the apple orchard. Six samples of apple trees, representing a variety of shapes and sizes, were selected for data collection and validation. Commercial software and the python programming language were utilized to process the collected data. The data processing steps involved data conversion, radius outlier removal, voxel grid downsampling, denoising through filtering and erroneous points, segmentation of the region of interest (ROI), clustering using the density-based spatial clustering (DBSCAN) algorithm, data transformation, and the removal of ground points. Accuracy was assessed by comparing the estimated outputs from the point cloud with the corresponding measured values. The sensor-estimated and measured tree heights were 3.05 ± 0.34 m and 3.13 ± 0.33 m, respectively, with a mean absolute error (MAE) of 0.08 m, a root mean squared error (RMSE) of 0.09 m, a linear coefficient of determination (r 2 ) of 0.98, a confidence interval (CI) of -0.14 to -0.02 m, and a high concordance correlation coefficient (CCC) of 0.96, indicating strong agreement and high accuracy. The sensor-estimated and measured canopy volumes were 13.76 ± 2.46 m 3 and 14.09 ± 2.10 m 3 , respectively, with an MAE of 0.57 m 3 , an RMSE of 0.61 m 3 , an r 2 value of 0.97, and a CI of -0.92 to 0.26, demonstrating high precision. For tree and row spacing, the sensor-estimated distances and measured distances were 3.04 ± 0.17 and 3.18 ± 0.24 m, and 3.35 ± 0.08 and 3.40 ± 0.05 m, respectively, with RMSE and r 2 values of 0.12 m and 0.92 for tree spacing, and 0.07 m and 0.94 for row spacing, respectively. The MAE and CI values were 0.09 m, 0.05 m, and -0.18 for tree spacing and 0.01, -0.1, and 0.002 for row spacing, respectively. Although minor differences were observed, the sensor estimates were efficient, though specific measurements require further refinement. The results are based on a limited dataset of six measured values, providing initial insights into geometric feature characterization performance. However, a larger dataset would offer a more reliable accuracy assessment. The small sample size (six apple trees) limits the generalizability of the findings and necessitates caution in interpreting the results. Future studies should incorporate a broader and more diverse dataset to validate and refine the characterization, enhancing management practices in apple orchards.

Why it matches plant phenotyping methods3D LiDARと点群処理によりリンゴ樹の樹高・樹冠体積・間隔を推定し、実測値と精度検証しており、植物形質取得法が研究の中心です。

abstractThis study aimed to quantify the height, canopy volume, tree spacing, and row spacing in an apple orchard using a three-dimensional (3D) LiDAR sensor.

A Comprehensive Review of Research on Disease Prediction in Plant Leaves →

In agriculture, plant disease detection at an early stage is significant. Plant leaves are an important factor in plant disease detection. Early disease prevention in-line harvest loss benefit is given to plant growers. The odd leaves can be visible after getting infected, and if the software can tell accurately the disease infestation result. Forecasting plant leaf disease works better for an early disease warning. The infected leaf needs to be permanently removed, or it will affect the plant. As leaves are a significant part and produce food from the sun, the infected leaves show different patterns. Many proposing such AI and ML models have forecasted the infected leaves of many plants. This document focuses on forecasting diseases in plant leaves mainly for four types of plants: tomato, potato, apple & grape. Diseases in plant leaves can cause significant damage to plants and can have a detrimental effect on both the quantity and quality of production whether we consider a lower scale kitchen gardening or a high scale agriculture farming. The research begins with an overview of the common plant leaf diseases found in each of these four plants, along with their symptoms and causes. It then discusses the importance of early detection and management of these diseases to ensure healthy growth. This helps a naïve gardener as well as an experienced farmer to prevent a plant or a whole field from getting infected or diseased. Plant disease detection approaches are crucial for prevention and management. Various techniques in plant disease detection using AI- based machine learning and deep learning methods are reviewed comprehensively. Keywords— Agriculture farming, Kitchen Garden, Plant leaf diseases, Forecasting diseases.

Why it matches plant phenotyping methods植物葉の病徴・病害をAI/機械学習で検出・予測する手法を包括的にレビューしており、植物の病害状態を観測・推定するフェノタイピング手法が中心です。

titleA Comprehensive Review of Research on Disease Prediction in Plant Leaves

Selection of volatile markers for rubbery rot in apple fruit caused by Phacidiopycnis washingtonensis. →

Although a major share of postharvest losses of apples is due to fungal fruit rots, their timely detection is difficult in commercial bulk-storage rooms. Therefore, a method was developed to identify the volatile markers of fruit naturally infected by Phacidiopycnis washingtonensis, a common storage-rot fungus of Northern Europe, and North and South America. Potato dextrose agar, apple juice agar, and fruit of the apple cultivar 'Nicoter' were inoculated with P. washingtonensis. Volatile organic compounds (VOCs) were sampled from the headspace of inoculated and uninoculated agar cultures and fruits using solid-phase micro-extraction and analysed by gas chromatography-mass spectrometry. The number of emitted alcohols and miscellaneous compounds was higher from agar and fruit colonised by P. washingtonensis than from uninoculated controls, whereas more aldehydes and esters were detected in uninoculated samples. These results indicate that the fungus produced alcohols and miscellaneous compounds and consumed aldehydes and esters while growing. The concentration of 37 of the VOCs was higher in the P. washingtonensis inoculated agar compared to the uninoculated agar, and nine of these compounds (3-methyl-1-butanol, 3-methyl-2-buten-1-ol, 2-phenylethanol, acetone, 3-methyl furan, styrene, 1-ethyl-4-methoxybenzene, 4-ethylphenol, and 4-ethyl-2-methoxyphenol) were associated with fungal growth both in vitro and in vivo. Twenty-nine compounds were also detected in higher concentrations in apple fruit naturally infected by P. washingtonensis, indicating that the VOC method has potential as an early warning of storage rot in apples.

Why it matches plant phenotyping methodsリンゴ果実の真菌感染状態を揮発性有機化合物で検出する方法を開発し、感染果実で検証している。病害状態の取得・早期検出が中心で、単なる生物学的測定ではない。

abstractTherefore, a method was developed to identify the volatile markers of fruit naturally infected by Phacidiopycnis washingtonensis

Detection of Apple Proliferation Disease Using Hyperspectral Imaging and Machine Learning Techniques. →

Apple proliferation is among the most important diseases in European fruit production. Early and reliable detection enables farmers to respond appropriately and to prevent further spreading of the disease. Traditional phenotyping approaches by human observers consider multiple symptoms, but these are difficult to measure automatically in the field. Therefore, the potential of hyperspectral imaging in combination with data analysis by machine learning algorithms was investigated to detect the symptoms solely based on the spectral signature of collected leaf samples. In the growing seasons 2019 and 2020, a total of 1160 leaf samples were collected. Hyperspectral imaging with a dual camera setup in spectral bands from 400 nm to 2500 nm was accompanied with subsequent PCR analysis of the samples to provide reference data for the machine learning approaches. Data processing consists of preprocessing for segmentation of the leaf area, feature extraction, classification and subsequent analysis of relevance of spectral bands. The results show that imaging multiple leaves of a tree enhances detection results, that spectral indices are a robust means to detect the diseased trees, and that the potentials of the full spectral range can be exploited using machine learning approaches. Classification models like rRBF achieved an accuracy of 0.971 in a controlled environment with stratified data for a single variety. Combined models for multiple varieties from field test samples achieved classification accuracies of 0.731. Including spatial distribution of spectral data further improves the results to 0.751. Prediction of qPCR results by regression based on spectral data achieved RMSE of 14.491 phytoplasma per plant cell.

Why it matches plant phenotyping methods葉のハイパースペクトル画像からリンゴ増殖病の症状・感染状態を抽出し、機械学習で分類・定量する手法が研究の中心であり、精度評価も実施しているため。

abstractthe potential of hyperspectral imaging in combination with data analysis by machine learning algorithms was investigated to detect the symptoms solely based on the spectral signature of collected leaf samples

Developments in deep learning approaches for apple leaf Alternaria disease identification: A review →

Apple tree leaf diseases (ATLDs) can be accurately identified and addressed early to prevent the diseases from spreading, minimize the need for chemical pesticides and fertilizers, increase apple quality and production, and preserve the healthy growth of apple varieties. To overcome such challenges, different Deep Learning (DL) approaches have been developed to early detect apple leaf diseases. In this paper, the data from 2010 to 2024 has been taken for analysis, and it has been observed that many of the researchers have utilized different types of datasets for disease detection. Moreover, Deep Learning (DL) and Machine Learning (ML) have been mostly utilized for the detection and identification of apple leaf Alternaria diseases. It has also been observed from the previous work that Support Vector Machines (SVM), Random Forests (RF), XGBoost, and many more are the most common approaches utilized by the researchers. On the other hand, DenseNet, MobileNet, Convolutional Neural Network (CNN), and Vision Transformer are the deep learning approaches utilized by the researchers. Furthermore, we have also given a brief analysis of each approach along with a comparative analysis such as lightweight CNNs and Attention-based mechanisms, Transfer Learning (TL), Localization techniques, Vision Transformer (ViT), and Severity estimation techniques. Emphasizing their methods, datasets, performance metrics, and real-world applications. This study explores the proposed models’ approaches, feature selection and extraction techniques, data capturing conditions, accuracy, types of datasets used in the experiments, and their resources. Our research findings indicate that although DL approaches have significant potential for improving disease management in agriculture. There is a crucial need for a more scalable, robust, and flexible solution to handle numerous agricultural conditions and disease complexities. By methodically and comprehensively analyzing the collected data, this study aims to facilitate valuable resources for researchers aiming to design, develop, and implement DL-based systems for apple leaf disease detection and identification, ultimately contributing to sustainable agriculture and improved food security.

Why it matches plant phenotyping methodsリンゴ葉の病害を画像から検出・同定し、重症度推定を含む深層学習手法、データセット、性能評価を体系的にレビューしており、植物病態のフェノタイピング手法が中心である。

titleDevelopments in deep learning approaches for apple leaf Alternaria disease identification: A review

Multi-scale and multi-receptive field-based feature fusion for robust segmentation of plant disease and fruit using agricultural images →

The accurate and fast assessment of plant diseases and fruits is important for sustainable and productive agriculture. However, typically manual methods are adopted for the assessment of plant diseases and fruit, which is a time-consuming, resource-intensive, and error-prone approach. A few artificial intelligence (AI)-based methods have been introduced to automate this process, but they show limitations in attaining high performance in challenging imaging conditions such as occlusions, poor lighting, noise, indistinctive boundaries, and excessive morphological variations. Moreover, current approaches also lack in providing computationally efficient solutions. To overcome these issues, two novel architectures are developed to segment plant diseases and fruits with higher segmentation performance and less computational requirements. The efficient feature fusion segmentation network (EFFS-Net) is the base network, and a multi-scale dilated feature fusion segmentation network (MDFS-Net) is the final network of this study. EFFS-Net uses an identity skip path-based feature fusion mechanism with an efficient grouped convolutional depth (EGCD) to provide satisfactory segmentation performance with high computational efficiency. MDFS-Net uses a multi-scale feature fusion mechanism and fuses low-level information in the EGCD and other sections of the network to learn detailed input features for delivering promising performance even in challenging imaging conditions. MDFS-Net also applies multiple receptive fields to the low-level information in the effective receptive field processing (ERFP) block for fusion near the pixel classification stage for further performance enhancement. Both networks are evaluated using a Brazilian Arabica coffee leaf (BRACOL) image dataset, an Australian Center for Field Robotics orchard fruit (apple) dataset, and a necrotized cassava root cross-section image dataset. The proposed method provides a promising segmentation performance, achieving dice similarity coefficients of 88.81 %, 95.01 %, and 86.04 % with performance improvement of 3.5 %, 2.6 %, and 1.07 % on the three datasets, respectively, with approximately ten times less number of required trainable parameters compared with the state-of-the-art methods. • Architectures (EFFS-Net and MDFS-Net) for plant disease and fruit segmentation. • EFFS-Net uses efficient grouped convolutional depth. • MDFS-Net uses multi-scale and multi-receptive field-based feature fusion mechanisms. • Our trained models and codes are publicly available via Github site.

Why it matches plant phenotyping methods植物病害および果実を画像からセグメンテーションする新規ネットワークを開発・評価しており、植物の病害状態や器官の抽出が中心的な方法論的貢献である。

abstracttwo novel architectures are developed to segment plant diseases and fruits with higher segmentation performance and less computational requirements

A pooling module with multidirectional and multi-scale spatial information and its application on semantic segmentation of leaf lesions →

Timely and accurate identification of apple leaf diseases provides an important basis for the early warning and precise control of apple leaf diseases. It was of great significance to reduce the economic losses caused by diseases. In order to improve the accuracy of apple leaf lesion segmentation by using the deep neural networks, this paper proposed a twill pooling method and combined it with strip pooling to propose the double cross pooling method. The pooling method could contract multidirectional and multi-scale spatial context information. Then, on the constructed apple leaf disease dataset, these modules were combined with existing deep semantic segmentation networks to perform disease lesion semantic segmentation. These three constructed modules were added to the fully convolutional network respectively, and the universality of the constructed modules was further verified on other semantic segmentation networks. Experimental results showed that the proposed modules could improve the mean intersection over union of fully convolutional network by 7.69%, up to 80.44% on the collected apple leaf disease dataset. Furthermore, when adding the proposed pooling modules to DeepLabV3 + , PSPNet and U-Net, the mean intersection over union improved by varying degrees when adding any of the proposed modules alone. The performance of the improved fully convolutional network, DeepLabV3 + , PSPNet and U-Net were compared. The DeepLabV3 + with the double cross pooling module had the best segmentation performance whose mean pixel accuracy, mean intersection over union, leaf disease classification accuracy and leaf disease degree diagnosis accuracy were 99.07%, 82.1%, 99.52% and 77.65% respectively.

Why it matches plant phenotyping methodsリンゴ葉病斑の画像セグメンテーション手法を開発し、複数のネットワークとデータセットで性能検証しているため、植物病害状態の画像ベース表現型計測手法が中心である。

abstractthis paper proposed a twill pooling method and combined it with strip pooling to propose the double cross pooling method

AgRegNet: A deep regression network for flower and fruit density estimation, localization, and counting in orchards →

One of the major challenges for the agricultural industry today is the uncertainty in manual labor availability and the associated cost. Automated flower and fruit density estimation, localization, and counting could help streamline harvesting, yield estimation, and crop-load management strategies such as flower and fruitlet thinning. This article proposes a deep regression-based network, AgRegNet, to estimate density, count, and location of flower and fruit in tree fruit canopies without explicit detection or polygon annotation. Inspired by popular U-Net architecture, AgRegNet is a U-shaped network with an encoder-to-decoder skip connection and modified ConvNeXt-T as an encoder feature extractor. AgRegNet can be trained based on information from point annotation and leverages segmentation information and attention modules (spatial and channel) to highlight relevant flower and fruit features while suppressing non-relevant background features. Experimental evaluation in apple flower and fruit canopy images under an unstructured orchard environment showed that AgRegNet achieved promising accuracy as measured by Structural Similarity Index (SSIM), percentage Mean Absolute Error (pMAE) and mean Average Precision (mAP) to estimate flower and fruit density, count, and centroid location, respectively. Specifically, the SSIM, pMAE, and mAP values for flower images were 0.938, 13.7%, and 0.81, respectively. For fruit images, the corresponding values were 0.910, 5.6%, and 0.93. Since the proposed approach relies on information from point annotation, it is suitable for sparsely and densely located objects. This simplified technique will be highly applicable for growers to accurately estimate yields and decide on optimal chemical and mechanical flower thinning practices.

Why it matches plant phenotyping methods花および果実の密度・個数・位置を画像から推定する深層学習手法を提案し、実画像で性能評価しており、植物表現型取得法が研究の中心である。

abstractThis article proposes a deep regression-based network, AgRegNet, to estimate density, count, and location of flower and fruit in tree fruit canopies without explicit detection or polygon annotation.

Semantic segmentation for plant leaf disease classification and damage detection: A deep learning approach →

Agriculture sustains the livelihoods of a significant portion of India's rural population, yet challenges persist in manual practices and disease management. To address these issues, this paper presents an automated plant leaf damage detection and disease identification system leveraging advanced deep learning techniques. The proposed method consists of six stages: first, utilizing YOLOv8 for region of interest identification from drone images; second, employing DeepLabV3+ for background removal and facilitating disease classification; third, implementing a CNN model for accurate disease classification achieving high training and validation accuracies (96.97 % and 92.89 %, respectively); fourth, utilizing UNet semantic segmentation for precise damage detection at a pixel level with an evaluation accuracy of 99 %; fifth, evaluating disease severity; and sixth, suggesting tailored remedies based on disease type and damage state. Experimental analysis using the Plant Village dataset demonstrates the effectiveness of the proposed method in detecting various defects in plants such as apple, tomato, and corn. This automated approach holds promise for enhancing agricultural productivity and disease management in India and beyond.

Why it matches plant phenotyping methods植物葉の病害・損傷を画像から分類、画素単位で検出し、病害重症度を評価する深層学習手法が研究の中心であるため、植物状態の画像ベース表現型解析として採用。

abstractthis paper presents an automated plant leaf damage detection and disease identification system leveraging advanced deep learning techniques.

An image-based system for locating pruning points in apple trees using instance segmentation and RGB-D images →

Intelligent pruning is an effective way to improve the efficiency of fruit tree pruning and reduce production costs, where the positioning of fruit tree pruning points is the key. In this study, a pruning point localisation method based on deep learning and red-green-blue-depth (RGB-D) is proposed for dormant tall spindle apple trees, which can quickly and accurately identify the pruning points on the primary branches. Firstly, red-green-blue (RGB) images and depth images of apple trees were acquired by using the Realsense D435i depth camera, and the SOLOv2 instance segmentation model was used to segment the trunks, branches, and supports in RGB images. Secondly, the manual pruning rules were adapted to improve the pruning methods; then using OpenCV image processing method, the junction points of branches and trunk and potential pruning points were gained, and the world coordinates of the points were obtained according to the coordinate transformation to calculate the length of branch diameter and spacing. Finally, the pruning points were determined according to the pruning rules. The results show that SOLOv2 has better segmentation effect compared with Mask R–CNN and Cascade Mask R–CNN. The mean absolute error between the estimated and manually measured values of branch diameter and spacing are 1.10 mm and 16.06 mm, and the recognition accuracy of pruning points is 87.2%, with a recognition time of about 3.8 s for each image. It is shown that the method can measure branch diameter and spacing and quickly locate pruning points with high reliability and accuracy, and the study provides a basis for the development of apple tree pruning robots.

Why it matches plant phenotyping methodsRGB-D画像と画像処理・深層学習による枝径・枝間隔の推定および剪定点抽出が研究の中心であり、植物形態形質の取得手法を開発・検証しているため。

abstracta pruning point localisation method based on deep learning and red-green-blue-depth (RGB-D) is proposed

Non-destructive detection of codling moth infestation in apples using acoustic impulse response signals →

Codling moth (CM) (Cydia pomonella L.) is the most destructive pest for apples, causing large economic losses when not properly mitigated. Efficient detection methods can limit the spread of this pest in the apple supply chain. Non-destructive methods have several advantages over the current methods in that they can be applied to every apple (or a much larger sample) thereby reducing the possibility of missed detection. This paper examines the feasibility of acoustic impulse response methods for detecting CM larvae-infested apples. Experiments were performed on control and artificially infested apples from three different cultivars. Signals were recorded with a contact sensor, and 21 signal features were proposed and extracted to characterise relevant properties of the response. The 21 features were evaluated with 11 machine leaning algorithms to determine if the features or their subsets contained information that could reliability determine if an apple was/is infested. Classification test results using a 10-fold cross-validation indicated accuracy rates between 80% and 92% for Fuji apples, between 92% and 99% for Gala apples, and 63% and 97% for Granny Smith apples. The impulse response required between 60 and 80 ms for each apple (not counting setup/transition time). These results from this study suggest that active impulse response classification can potentially improve the detection of post-harvest apple CM infestation detection along the supply chain.

Why it matches plant phenotyping methodsリンゴ果実の害虫幼虫による感染・食害状態を、音響インパルス応答と特徴量・機械学習で非破壊推定する方法が研究の中心であり、植物器官の状態推定に該当する。

abstractThis paper examines the feasibility of acoustic impulse response methods for detecting CM larvae-infested apples.

Plant Leaf Disease Detection Using Deep Learning: A Multi-Dataset Approach →

This paper introduces a deep learning approach for detecting plant leaf diseases. The objective is to develop robust models capable of accurately identifying plant diseases across various image backgrounds, thereby overcoming the limitations of existing methods that often rely on controlled laboratory conditions. To achieve this, a combination of the PlantDoc dataset and Web-sourced data of plant images from online platforms was used. This paper implemented and compared state-of-the-art *cnn architectures, including EfficientNet-B0, EfficientNet-B3, ResNet50, and DenseNet201, all fine-tuned specifically for leaf disease classification. A significant contribution is the application of enhanced data augmentation techniques, such as adding Gaussian noise, to improve model generalisation. Results indicated varied performance across the datasets, with EfficientNet models generally outperforming others. When trained and tested on the PlantDoc dataset, EfficientNet-B3 achieved the highest accuracy of 73.31%. In cross-dataset evaluation, EfficientNet-B3 reached 76.77% accuracy when trained on PlantDoc and tested on the Web-sourced dataset. The best performance occurred when training on the combined dataset and testing on the Web-sourced data, resulting in an accuracy of 80.19%. Class-wise F1-scores revealed consistently high performance (>0.90) for diseases such as apple rust leaf and grape leaf across models. This paper contributes to the comparative analysis of various datasets and model architectures for effective leaf disease detection.

Why it matches plant phenotyping methods植物葉の病害状態を画像から推定する深層学習手法の開発・比較が研究の中心であり、データセット間評価も実施しているため、植物フェノタイピング手法として含める。

abstractThis paper introduces a deep learning approach for detecting plant leaf diseases.

RGB-D Camera and Fractal-Geometry-Based Maximum Diameter Estimation Method of Apples for Robot Intelligent Selective Graded Harvesting →

Realizing the integration of intelligent fruit picking and grading for apple harvesting robots is an inevitable requirement for the future development of smart agriculture and precision agriculture. Therefore, an apple maximum diameter estimation model based on RGB-D camera fusion depth information was proposed in the study. Firstly, the maximum diameter parameters of Red Fuji apples were collected, and the results were statistically analyzed. Then, based on the Intel RealSense D435 RGB-D depth camera and LabelImg software, the depth information of apples and the two-dimensional size information of fruit images were obtained. Furthermore, the relationship between fruit depth information, two-dimensional size information of fruit images, and the maximum diameter of apples was explored. Based on Origin software, multiple regression analysis and nonlinear surface fitting were used to analyze the correlation between fruit depth, diagonal length of fruit bounding rectangle, and maximum diameter. A model for estimating the maximum diameter of apples was constructed. Finally, the constructed maximum diameter estimation model was experimentally validated and evaluated for imitation apples in the laboratory and fruits on the Red Fuji fruit trees in modern apple orchards. The experimental results showed that the average maximum relative error of the constructed model in the laboratory imitation apple validation set was ±4.1%, the correlation coefficient (R2) of the estimated model was 0.98613, and the root mean square error (RMSE) was 3.21 mm. The average maximum diameter estimation relative error on the modern orchard Red Fuji apple validation set was ±3.77%, the correlation coefficient (R2) of the estimation model was 0.84, and the root mean square error (RMSE) was 3.95 mm. The proposed model can provide theoretical basis and technical support for the selective apple-picking operation of intelligent robots based on apple size grading.

Why it matches plant phenotyping methodsRGB-D画像と深度情報からリンゴ果実の最大径を推定する手法を開発し、実験室および果樹園で検証しており、収穫ロボット用途を超えて果実形質の取得が中心である。

abstractan apple maximum diameter estimation model based on RGB-D camera fusion depth information was proposed in the study

Detection of Apple Sucrose Concentration Based on Fluorescence Hyperspectral Image System and Machine Learning. →

China ranks first in apple production worldwide, making the assessment of apple quality a critical factor in agriculture. Sucrose concentration (SC) is a key factor influencing the flavor and ripeness of apples, serving as an important quality indicator. Nondestructive SC detection has significant practical value. Currently, SC is mainly measured using handheld refractometers, hydrometers, electronic tongues, and saccharimeter analyses, which are not only time-consuming and labor-intensive but also destructive to the sample. Therefore, a rapid nondestructive method is essential. The fluorescence hyperspectral imaging system (FHIS) is a tool for nondestructive detection. Upon excitation by the fluorescent light source, apples displayed distinct fluorescence characteristics within the 440-530 nm and 680-780 nm wavelength ranges, enabling the FHIS to detect SC. This study used FHIS combined with machine learning (ML) to predict SC at the apple's equatorial position. Primary features were extracted using variable importance projection (VIP), the successive projection algorithm (SPA), and extreme gradient boosting (XGBoost). Secondary feature extraction was also conducted. Models like gradient boosting decision tree (GBDT), random forest (RF), and LightGBM were used to predict SC. VN-SPA + VIP-LightGBM achieved the highest accuracy, with Rp2, RMSEp, and RPD reaching 0.9074, 0.4656, and 3.2877, respectively. These results underscore the efficacy of FHIS in predicting apple SC, highlighting its potential for application in nondestructive quality assessment within the agricultural sector.

Why it matches plant phenotyping methodsリンゴ果実の糖濃度という植物器官形質を、蛍光ハイパースペクトル画像と機械学習で非破壊推定する手法が研究の中心であり、モデル比較・特徴抽出・精度評価も行っている。

abstractThis study used FHIS combined with machine learning (ML) to predict SC at the apple's equatorial position.

Non-destructive assessment of apple internal quality using rotational hyperspectral imaging. →

This work aims to predict the starch, vitamin C, soluble solids, and titratable acid contents of apple fruits using hyperspectral imaging combined with machine learning approaches. First, a hyperspectral camera by rotating samples was used to obtain hyperspectral images of the apple fruit surface in the spectral range of 380~1018 nm, and its region of interest (ROI) was extracted; then, the optimal preprocessing method was preferred through experimental comparisons; on this basis, genetic algorithms (GA), successive projection algorithms (SPA), and competitive adaptive reweighting adoption algorithms (CARS) were used to extract feature variables; subsequently, multiple machine learning models (support vector regression SVR, principal component regression PCR, partial least squares regression PLSR, and multiple linear regression MLR) were used to model the inversion between hyperspectral images and internal nutrient quality physicochemical indexes of fruits, respectively. Through the comparative analysis of all the model prediction results, it was found that among them, for starch, vitamin C, soluble solids, and titratable acid content, 2 nd Der-CARS-MLR were the optimal prediction models with superior performance (the prediction coefficients of determination R p 2 exceeded 90% in all of them). In addition, potential relationships among four nutritional qualities were explored based on t-values and p-values, and a significant conclusion was drew that starch and vitamin C was highly correlated.

Why it matches plant phenotyping methods回転ハイパースペクトル画像と機械学習により、リンゴ果実の内部品質形質を非破壊推定する手法の開発・比較が中心である。

titleNon-destructive assessment of apple internal quality using rotational hyperspectral imaging.

Detection of early bruises in apples using hyperspectral imaging and an improved MobileViT network. →

Apples are susceptible to postharvest bruises, leading to a shortened shelf life and significant waste. Therefore, accurate detection of apple bruises is crucial to mitigate food waste. This study proposed an improved lightweight network based on MobileViT for detecting early-stage bruises in apples, utilizing hyperspectral imaging technology from 397.66 to 1003.81 nm. After acquiring hyperspectral images, the Otsu threshold algorithm was employed for mask extraction, and principal component analysis was used for feature image extraction. Subsequently, the improved MobileViT network (iM-ViT) was implemented and compared with traditional algorithms, utilizing depthwise separable convolutions for parameter reduction and integrating local and global features to enhance bruise detection capability. The results demonstrated the superior performance of iM-ViT in accurately detecting apple bruises, showing significant improvements. The F1 score and test accuracy for detecting apple bruises using iM-ViT reached 0.99 and 99.07%, respectively. The fivefold cross-validation strategy was used to assess the stability and robustness of iM-ViT, and ablation experiments were performed to explore the effects of depthwise separable convolutions and local features on parameter reduction and classification accuracy improvement for early-stage bruise detection in apples. The results demonstrated that iM-ViT effectively reduced parameters and improved the ability to detect early bruises in apples. PRACTICAL APPLICATION: This study proposed an improved lightweight network to detect early bruises in apples, providing a reference for quick detection of bruises caused in the production process. Potential insights into the nondestructive detection of apple bruises using lightweight networks have been presented, which might be applied to mobile or online devices.

Why it matches plant phenotyping methodsリンゴ果実の打撲という植物器官の状態を、ハイパースペクトル画像と改良MobileViTで非破壊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す。

abstractThis study proposed an improved lightweight network based on MobileViT for detecting early-stage bruises in apples, utilizing hyperspectral imaging technology from 397.66 to 1003.81 nm.

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

Incremental RPN: Hierarchical Region Proposal Network for Apple Leaf Disease Detection in Natural Environments. →

Apple leaf diseases can seriously affect apple production and quality, and accurately detecting them can improve the efficiency of disease monitoring. Owing to the complex natural growth environment, apple leaf lesions may be easily confused with background noise, leading to poor performance. In this study, a cascaded Incremental Region Proposal Network (Inc-RPN) is proposed to accurately detect apple leaf diseases in natural environments. The proposed Inc-RPN has a two-layer RPN architecture, where the precursor RPN is leveraged to generate diseased leaf proposals, and the successor RPN focuses on extracting target disease spots based on diseased leaf proposals. In the successor RPN, a low-level feature aggregation module is designed to fully utilize the bridged features and preserve the semantic information of the target disease spots. An incremental module is also leveraged to extract aggregated diseased leaf features and target disease spot features. Finally, a novel position anchor generator is designed to generate anchors based on diseased leaf proposals. The experimental results show that the proposed Inc-RPN performs very well on the FALD_CED and Apple Leaf Disease datasets, showing that it can accurately perform apple leaf disease detection tasks.

Why it matches plant phenotyping methodsリンゴ葉の病斑・病害を画像から検出する新規深層学習手法を開発し、複数データセットで性能評価しており、植物の病害状態を取得する方法が中心である。

abstractIn this study, a cascaded Incremental Region Proposal Network (Inc-RPN) is proposed to accurately detect apple leaf diseases in natural environments.

Evaluation of a hyperspectral image pipeline toward building a generalisation capable crop dry matter content prediction model →

Hyperspectral imaging has proven to be a reliable technique for estimating dry matter, a common variable when considering the quality of the fresh produce. However, developing models capable of generalising across different crops is challenging. In this study, several pipelines were explored towards achieving a robust and accurate generic regression model were evaluated and the development of Automatic Relevance Determination (ARD) and Partial Least Squares (PLS) algorithms for fruit and vegetable dry matter estimation. The models were built using a VIS-NIR dataset that includes both fruit and vegetables, namely, apples, broccoli and leek (n = 779). The PLS regression model obtained Root Mean Square on Prediction (RMSEP) = 0.0137, outperforming ARD regression (RMSEP = 0.0140) on a 10x5-fold cross-validation protocol. The evaluated preprocessing techniques affect the two regression algorithms differently, with the best results achieved when the pipeline was used without feature extraction. Overall, the pipeline using either ARD or PLS regression shows strong performance and generalisation for Visible-Near Infrared (VIS-NIR)-based dry matter estimation across diverse fruits and vegetables.

Why it matches plant phenotyping methodsVIS-NIRハイパースペクトル画像から果実・野菜の乾物含量を推定するパイプラインと回帰モデルを比較・検証しており、植物形質取得が研究の中心である。

titleEvaluation of a hyperspectral image pipeline toward building a generalisation capable crop dry matter content prediction model

Automatic Apple Detection and Counting with AD-YOLO and MR-SORT. →

In the production management of agriculture, accurate fruit counting plays a vital role in the orchard yield estimation and appropriate production decisions. Although recent tracking-by-detection algorithms have emerged as a promising fruit-counting method, they still cannot completely avoid fruit occlusion and light variations in complex orchard environments, and it is difficult to realize automatic and accurate apple counting. In this paper, a video-based multiple-object tracking method, MR-SORT (Multiple Rematching SORT), is proposed based on the improved YOLOv8 and BoT-SORT. First, we propose the AD-YOLO model, which aims to reduce the number of incorrect detections during object tracking. In the YOLOv8s backbone network, an Omni-dimensional Dynamic Convolution (ODConv) module is used to extract local feature information and enhance the model's ability better; a Global Attention Mechanism (GAM) is introduced to improve the detection ability of a foreground object (apple) in the whole image; a Soft Spatial Pyramid Pooling Layer (SSPPL) is designed to reduce the feature information dispersion and increase the sensory field of the network. Then, the improved BoT-SORT algorithm is proposed by fusing the verification mechanism, SURF feature descriptors, and the Vector of Local Aggregate Descriptors (VLAD) algorithm, which can match apples more accurately in adjacent video frames and reduce the probability of ID switching in the tracking process. The results show that the mAP metrics of the proposed AD-YOLO model are 3.1% higher than those of the YOLOv8 model, reaching 96.4%. The improved tracking algorithm has 297 fewer ID switches, which is 35.6% less than the original algorithm. The multiple-object tracking accuracy of the improved algorithm reached 85.6%, and the average counting error was reduced to 0.07. The coefficient of determination R2 between the ground truth and the predicted value reached 0.98. The above metrics show that our method can give more accurate counting results for apples and even other types of fruit.

Why it matches plant phenotyping methodsリンゴ果実の検出・追跡・計数手法を開発し、精度や計数誤差を検証しており、果実数という植物形質の取得が中心である。

abstracta video-based multiple-object tracking method, MR-SORT (Multiple Rematching SORT), is proposed

YOLO11 and Vision Transformers based 3D Pose Estimation of Immature Green Fruits in Commercial Apple Orchards for Robotic Thinning →

In this study, a robust method for 3D pose estimation of immature green apples (fruitlets) in commercial orchards was developed, utilizing the YOLO11 object pose detection model alongside Vision Transformers (ViT) for depth estimation. For object detection and pose estimation, performance comparisons of YOLO11 (YOLO11n, YOLO11s, YOLO11m, YOLO11l and YOLO11x) and YOLOv8 (YOLOv8n, YOLOv8s, YOLOv8m, YOLOv8l and YOLOv8x) were made under identical hyperparameter settings among the all configurations. Likewise, for RGB to RGB-D mapping, Dense Prediction Transformer (DPT) and Depth Anything V2 were investigated. It was observed that YOLO11n surpassed all configurations of YOLO11 and YOLOv8 in terms of box precision and pose precision, achieving scores of 0.91 and 0.915, respectively. Conversely, YOLOv8n exhibited the highest box and pose recall scores of 0.905 and 0.925, respectively. Regarding the mean average precision at 50% intersection over union (mAP@50), YOLO11s led all configurations with a box mAP@50 score of 0.94, while YOLOv8n achieved the highest pose mAP@50 score of 0.96. In terms of image processing speed, YOLO11n outperformed all configurations with an impressive inference speed of 2.7 ms, significantly faster than the quickest YOLOv8 configuration, YOLOv8n, which processed images in 7.8 ms. This demonstrates a substantial improvement in inference speed over previous iterations, particularly evident when comparing YOLO11n and YOLOv8n. Subsequent integration of ViTs for the green fruit's pose depth estimation revealed that Depth Anything V2 outperformed Dense Prediction Transformer in 3D pose length validation, achieving the lowest Root Mean Square Error (RMSE) of 1.52 and Mean Absolute Error (MAE) of 1.28, demonstrating exceptional precision in estimating immature green fruit lengths. Following this, the DPT showed notable accuracy improvements with a RMSE of 3.29 and an MAE of 2.62. In contrast, measurements derived from Intel RealSense point clouds exhibited the highest discrepancies from the ground truth, with a RMSE of 9.98 and an MAE of 7.74. These findings emphasize the effectiveness of YOLO11 in detecting and estimating the pose of immature green fruits, illustrating how Vision Transformers like Depth Anything V2 adeptly convert RGB images into RGB-D data, thus enhancing the precision and computational requirement of 3D pose estimations for future robotic thinning applications in commercial orchards.

Why it matches plant phenotyping methods未熟果実の検出にとどまらず、3D姿勢と果実長を推定し、複数モデルを比較・検証する画像ベースの植物器官形質計測法が中心である。

abstracta robust method for 3D pose estimation of immature green apples (fruitlets) in commercial orchards was developed

Detection of Apple Plant Diseases using Leaf Images through Convolution Neural Networks →

Effective identification and classification of plant diseases are critical for maintaining agricultural productivity and ensuring food security. With recent advancements in computer vision, Convolutional Neural Networks (CNNs) have emerged as powerful tools for diagnosing crop diseases. This research presents a novel approach for detecting apple plant diseases through the analysis of leaf images using CNNs. We utilize the PlantVillage dataset, which comprises a diverse collection of healthy and diseased apple leaf images. The dataset undergoes rigorous preprocessing to enhance image quality, followed by training and evaluation of the CNN models. Our approach achieves remarkable results, demonstrating an accuracy of 99.67%, precision of 98.23%, and F1 score of 99.04%. These findings affirm the robustness of our CNN-based model in accurately classifying various apple diseases. This research not only provides a reliable method for early disease detection but also emphasizes its significance in optimizing crop yield and safeguarding food security. By facilitating timely interventions, our proposed framework has the potential to mitigate the impact of diseases in apple orchards and can be extended to broader applications in precision agriculture.

Why it matches plant phenotyping methodsリンゴ葉画像から病害状態をCNNで推定する画像ベースの植物表現型解析手法が研究の中心であり、学習・評価による技術検証も行っている。

abstractThis research presents a novel approach for detecting apple plant diseases through the analysis of leaf images using CNNs.

A Comprehensive Hybrid Model For Apple Fruit Disease Detection Using Multi-Architecture Feature Extraction →

The agriculture industry is critical to the global economy, with product quality having a direct impact on marketability and waste management. Apples, one of the most extensively produced fruits, are affected by a variety of diseases that can reduce productivity and quality. Accurate diagnosis of diseases is critical, but traditional manual approaches are time-consuming, error-prone, and ineffective. Inadequate labeled data and a wide range of disease symptoms make it necessary to design an automated, robust, and accurate system. This article describes a hybrid model for Apple Fruit Disease Detection (HMAFDD) that combines the strengths of three pre-trained convolutional neural network (CNN) models: ResNet50, DenseNet121, and EfficientNetB0. It accomplish this by using multi-architecture feature extraction. The hybrid system, which combines both models, is able to recognize a wide range of features, from simple textures to complex patterns unique to a certain diseases. Grad-CAM, or gradient-weighted class activation mapping, creates heatmaps that highlight significant regions for prediction, which enhances the interpretability of the model.To increase robustness and accuracy, techniques such as spectral-shifted adversarial perturbation for data augmentation and spectrally-weighted global average pooling for feature aggregation are used. This technique provides 99.75% accuracy with minimal processing needs, making it acceptable for real-time applications in agricultural situations. This considerably improves apple disease management.

Why it matches plant phenotyping methodsリンゴ果実の病徴を画像から検出するCNNベースの手法開発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対象範囲に含まれる。

abstractThis article describes a hybrid model for Apple Fruit Disease Detection (HMAFDD) that combines the strengths of three pre-trained convolutional neural network (CNN) models: ResNet50, DenseNet121, and EfficientNetB0.

Feasibility and potential of terahertz spectral and imaging technology for Apple Valsa canker detection: A preliminary investigation. →

Apple Valsa canker (AVC) caused by the Ascomycete Valsa mali, seriously constrains the production and quality of apple fruits. The symptomless incubation characteristics of Valsa mali make it highly challenging to detect AVC at an early infection stage. After infecting the wound of apple bark, the pathogenic hyphae of AVC will expand and colonize the phloem tissue. Meanwhile, various enzymes and toxic substances released by hyphae cause the decomposition of cellulose and lignin, and the generation of poisonous secondary metabolites in bark tissue. However, these early symptoms of AVC are invisible from the bark's appearance. Fortunately, Terahertz Spectral Imaging (ThzSI) technology with the advantage of penetrating, and fingerprinting is promising for detecting hidden or slight symptoms of the fungal infection. This study is a preliminary investigation of terahertz frequency-domain spectra for AVC in the early stage of infection. Healthy and two-week-infected apple tree branches were prepared for capturing ThzS images, and the spectral data were preprocessed by Multivariate scattering correction (MSC), Savitzky-Golay convolution smoothing (SG), and standard normal variate (SNV) respectively to remove data noise and improve data quality. Principal component analysis (PCA), competitive adaptive reweighted sampling (CARS), and random frog (RFROG) were employed to extract the spectral feature bands to eliminate redundant data and improve computational efficiency. Machine learning models were established based on the spectral features to detect AVC at an early infection stage, where 11 of them exhibited the best performance with F1-score of 99.72%. To further explore disease information in spatial spectra, imaging data were acquired using terahertz imaging technology. Based on imaging data, pseudo-color imaging, histogram equalization, and Otsu segmentation were employed to visualize early infection areas in apple barks. Furthermore, histogram feature (HF), shape feature (SF), and local binary pattern (LBP) extracted from terahertz spectral images were utilized to establish the SVM, RF, and KNN models. HF-SF-KNN and HF-SF-LBP-KNN with the best performance achieved F1-score of 98.82%. This study presents a preliminary application of terahertz spectral and imaging technology for early-stage AVC detection and demonstrates its feasibility. Additionally, it provides a new way to detect AVC, which expands the application of ThzSI technology in tree disease detection in orchards and lays the foundation for further research.

Why it matches plant phenotyping methodsリンゴ樹皮の感染症状をテラヘルツ分光・画像から検出・可視化し、前処理、特徴抽出、画像分割、機械学習モデルを評価しているため、植物病害状態の取得・推定手法が中心である。

abstractTerahertz Spectral Imaging (ThzSI) technology with the advantage of penetrating, and fingerprinting is promising for detecting hidden or slight symptoms of the fungal infection.

Performance Analysis of Machine Learning Techniques in Plant Leaf Disease Detection →