Thermal imaging enables non-invasive assessment of canopy temperature, an essential indicator of plant stress, yet the lack of color cues and strong shadow interference make plant segmentation in thermal images difficult. Recent advances in foundation models have demonstrated improved performance and generalizability across applications, showing promise for domain-specific applications with limited annotated datasets such as plant segmentation in thermal images. This study investigates large multimodal models (LMMs) for thermal image segmentation in plant phenotyping. Building upon the SAM-CLIP framework, we design a unified pipeline spanning zero-shot inference, few-shot and low-shot fine-tuning, and active learning to maximize accuracy with minimal supervision. Evaluations on two thermal datasets, LadyBird Brassica and UGA Brassica, demonstrate robust performance after minimal adaptation across both datasets and superior performance compared with baselines, achieving mIoU D values of 97.54% on the LadyBird dataset and 76.94 % on the UGA dataset. We also release the resulting thermal segmentation annotations to support community benchmarking and reproducible research, highlighting the potential of LMMs to enable scalable, high-quality dataset construction for field phenotyping. The released datasets can be found at: https://cornell.box.com/s/dh69xf84464yrc1vlws92l1tflx7qa89
Why it matches plant phenotyping methods熱画像から植物を分割する手法を開発・評価し、植物フェノタイピング用データセットとアノテーションも公開しているため、フェノタイピング手法が中心的である。
abstractThis study investigates large multimodal models (LMMs) for thermal image segmentation in plant phenotyping.
Reproduction assets foundThe authors publicly released the paper-specific thermal segmentation annotations (20,538 LadyBird masks and 37,790 UGA masks) via a Cornell Box link stated in the abstract, results, and data availability statement. No author analysis code or trained model checkpoints are explicitly released; the mmsegmentation GitHub/Dataset · publicwe generated and publicly released segmentation annotations for the complete LadyBird and UGA thermal image datasets using the best-performing SAM-CLIP model. Specifically, the final model obtained through the multi-round training process was used to generate 20,538 masks for the LadyBird dataset and 37,790 masks for the UGA dataset. Details of the generated annotations are provided in Supplementary Fig. S1 , and both annotated datasets are publicly available at: https://cornell.box.com/s/dh69xf84464yrc1vlws92l1tflx7qa89Open asset ↗lines:220-232Code / dataset availability confirmedOpenAlex · Europe PMC · Crossref · checked 15 Sept 2026
Climate change threatens global Chinese cabbage ( Brassica rapa L. ssp. pekinensis ) production, a cool-season crop essential for Asian markets. With optimal growth at 18-20°C and severe disruption above 25°C, developing heat-resilient varieties is critical. This study integrated high-throughput 3D multispectral phenotyping with multivariate analysis to characterize temporal heat stress responses in 18 Chinese cabbage genotypes. Seedlings were subjected to heat stress (setpoint 40/35°C day/night; measured 35.7/31.5°C day/night air temperature) or controls (setpoint 25/20°C day/night; measured 25.0/17.7°C day/night air temperature) for 14 days, with continuous non-destructive monitoring of 14 morphological and spectral parameters using PlantEye F600 multispectral 3D scanner. Principal component analysis of temporal phenotyping data explained 62-68% of variance, enabling quantitative assessment of phenotypic stability through Euclidean distance measurements in PC space. Temporal analysis revealed crop-specific response patterns with maximum treatment separation at 3 days after treatment (DAT) (ΔC=3.27), reflecting Chinese cabbage’s rapid heat sensitivity as a cool-season crop, followed by progressive acclimation by 14 DAT (ΔC=1.41). Early responses (3-5 DAT) were dominated by morphological parameters, transitioning to physiological adjustments (10-14 DAT) characterized by spectral indices. Under heat stress, plants prioritized evaporative cooling through increased transpiration (four-fold increase) over carbon assimilation. A critical finding was the disproportionately greater reduction in root biomass relative to shoot biomass under to heat stress, with root biomass declining 38-47% versus 20% in shoots. Strong correlations (r>0.8) between 3D imaging parameters and destructive biomass measurements validated the non-destructive approach’s reliability. Notably, image-based root surface area analysis correlated strongly with actual root biomass (R 2 =0.698, p<0.001), enabling practical assessment of root area without conventional destructive processing. Based on integration of phenotypic stability (Euclidean distances in PC space) and biomass production under heat stress, this approach identified four distinct heat tolerance strategies: stable-productive genotypes (ideal breeding targets combining phenotypic stability with high heat-stress biomass production), stable-conservative genotypes (phenotypic stability with lower production), plastic-productive genotypes (substantial phenotypic changes yet high biomass production), and plastic-sensitive genotypes (phenotypically unstable and poor biomass production). This validated framework accelerates heat-tolerant Chinese cabbage breeding through efficient high-throughput phenotyping, enabling targeted genotype selection for diverse production environments facing climate warming.
Why it matches plant phenotyping methods3Dマルチスペクトルスキャナによる非破壊・時系列表現型取得と、その解析・検証が研究の中心であり、熱ストレス下の形態・生理形質を定量化する実質的なハイスループット表現型解析研究である。
abstractThis study integrated high-throughput 3D multispectral phenotyping with multivariate analysis to characterize temporal heat stress responses in 18 Chinese cabbage genotypes.
Reproduction assets foundThe paper states its collected phenotyping data are available in the supplementary material hosted with the article (open access under CC BY-NC-ND), making the paper-specific phenotype dataset publicly actionable via the article DOI. The analysis code, however, is only available from the corresponding author uponReasonDataset · publichrough field phenotyping.) between RDA and the World Vegetable Center (WorldVeg)” and by the long-term strategic donors to the WorldVeg: Taiwan, the United States, Australia, the United Kingdom, Germany, Thailand, South Korea, Philippines, and Japan.
Footnotes
Appendix A
Supplementary data to this article can be found online at https://doi.org/10.1016/j.plaphe.2026.100221 .
Appendix A.
Supplementary data
The following is the Supplementary data to this article.
Multimedia component 1
Data availability
The data collected and used in this study are available in the supplementary material. The code used for analysis can be obtained from the corresponding author upon reasonable request.
ReferenceOpen asset ↗lines:486-514Code / dataset availability confirmedEurope PMC · checked 5 Sept 2026
Monitoring the growth dynamics in field-grown cabbage is critically important for ensuring stable vegetable production and advancing precision agricultural management. However, conventional two-dimensional (2D) image-based monitoring approaches are limited to planar projection information and lack representations of spatial structural characteristics, rendering them inadequate for supporting high-precision, full-cycle phenotypic monitoring of cabbage under open-field conditions. In this study, a high-precision three-dimensional (3D) point cloud dataset covering the period from the seedling stage to maturity was constructed using depth cameras in conjunction with multi-view spatial registration techniques. Building on this dataset, an adaptive point cloud segmentation network designed for the whole-cycle growth monitoring was proposed, incorporating a Head Refinement Module (HRM), a Leaf Instance Segmentation Module (LISM), and Cross Module Interaction (CMI) to address leaf adhesion and head boundary delineation. Experimental results demonstrated that the proposed method consistently outperformed state-of-the-art models in both semantic and instance segmentation tasks. For semantic segmentation, the mean Intersection over Union (mIoU) reached 0.767, with a point classification accuracy of 94.8%. The model comprises 54.25 million parameters and achieves an average response time of 0.76 s. For instance segmentation, the Average Precision (AP) improved by 2.3% for cabbage heads and 3.8% for leaves, while the Average Recall (AR) increased by 6.9%. Growth parameters, including plant height and canopy spread, extracted from the segmentation results showed strong agreement with ground-truth measurements, with correlation of coefficients (R 2 ) exceeding 0.9 for plant height, canopy length, and canopy width. Leveraging these multidimensional phenotypic descriptors, the temporal dynamics of cabbage growth throughout the entire growth cycle were systematically characterized. Overall, this study enables dynamic monitoring of cabbage phenotypes across the full growth cycle, providing a novel technical pathway for extending 3D phenotyping from controlled environments to open-field applications and offering important support for precise crop monitoring and the development of digital twin agriculture.
Why it matches plant phenotyping methods3D点群データセット、セグメンテーションネットワーク、形質抽出を開発・検証し、圃場キャベツの草高や冠幅を定量化する植物フェノタイピング手法が研究の中心である。
abstracta high-precision three-dimensional (3D) point cloud dataset covering the period from the seedling stage to maturity was constructed using depth cameras in conjunction with multi-view spatial registration techniques.
Reproduction assets foundThe paper's authors publicly release their improved OneFormer3D point cloud segmentation code on GitHub; the cabbage 3D point cloud dataset is only available upon request.Code · publicThe code is available at https://github.com/PandaDalin/improve_oneformer3d. The data of this study are available from the corresponding author upon request.Open asset ↗PandaDalin/improve_oneformer3dhtml-lines:449-475Code / dataset availability confirmedEurope PMC · checked 5 Sept 2026
This data article describes a curated RGB-Depth image dataset captured using an Intel RealSense D435 stereo depth camera mounted on an autonomous mobile platform during field deployments at commercial baby broccoli farms in Victoria, Australia. The dataset comprises 1759 paired RGB images (640 × 480 pixels) and corresponding 16-bit depth frames acquired under both daytime (natural sunlight) and night-time (LED illumination) conditions, designed to support research in agricultural computer vision and robotic harvesting. Images were selected from 39,765 raw acquisitions through a reproducible Python curation pipeline applying quality filtering (blur detection, brightness thresholds, corruption detection), perceptual hash-based duplicate removal, and manual review. The final dataset includes 924 daytime and 835 night-time image pairs containing baby broccoli plants at various growth stages. The dataset provides RGB camera intrinsic parameters and pixel-aligned depth maps to enable 3D point cloud reconstruction. Potential applications include developing deep learning models for crop detection and segmentation, validating depth-based size estimation methods, and benchmarking illumination-robust vision systems. All data and curation code are publicly available under a CC BY 4.0 license.
Why it matches plant phenotyping methodsRGB-Depth画像データセットの構築と再現可能なキュレーションを中心とし、作物検出に加えてサイズ推定という植物形質の評価・ベンチマークに利用できるため。
titleA field-acquired RGB-Depth image dataset for computer vision-based baby broccoli detection and size estimation under varying illumination conditions.
Reproduction assets foundThe paper is a data article describing a public Mendeley Data repository containing the authors' field-acquired RGB-D baby broccoli image dataset (1759 image pairs, ground truth diameter annotations, camera intrinsics, and curation/annotation code), directly reproducing the paper's phenotyping measurements and analysisDataset · publicRepository name: Mendeley Data
Data identification number: 10.17632/px5p6zdk6k.3
Direct URL to data: https://data.mendeley.com/datasets/px5p6zdk6k/3Open asset ↗Mendeley Data · 10.17632/px5p6zdk6k.3html-lines:95-155Code / dataset availability confirmedEurope PMC · checked 14 Sept 2026
Cauliflower is among the more well-known vegetables there are. Consumed all around the globe due to it being rich in nutrients such as vitamins, antioxidants, and for being high in fibre. These are nutritional qualities that help with digestion, immune-system, and minimizing inflammation. It is a common issue among farmers to have to deal with various diseases in cauliflower leaves that are difficult to diagnose in their early stages. These diseases have a tendency to propagate in a really swift pace throughout entire fields worth of crops. This in-turn causes heavy losses in the harvest, and makes it much more tedious and resource-intensive to protect the crops. As a result, farmers get more likely to use high amounts of pesticides and harmful chemicals to streamline the process of getting a more reliable yield on their crops. This is not only costly, but it is also harmful both to the quality of crops and to the well-being of the environment. In this publication, we are introducing a dataset containing a considerable number of images of cauliflower leaves. This is intended to drive development on this topic at a faster pace than it is now, and to help enhance disease monitoring, diagnosis, and precautionary techniques. We collected our dataset images between November 2024 and January 2025. In this dataset, cauliflower leaves were categorized into three classes: Healthy, Insect Holes, and Black Rot, each reflecting a specific condition that impacts plant health at different stages. This dataset consists of 2,661 images. The pictures were captured at different locations in Bangladesh, under different weather conditions, dates, temperatures, and with different devices. To enhance the data quality, we used several steps to process the dataset, making sure it would reflect real-world conditions and be ready for training. The images were resized to a standard size of 3000 × 3000 pixels, brightness was adjusted to make the images more easily discernible, and we removed duplicates and poor-quality images. These actions helped ensure the dataset was in the best possible shape for effective model training. This dataset will be highly effective for agricultural research, precision agriculture, and effective management of diseases. It should help develop highly accurate machine learning models for early detection of Cauliflower leaf diseases. The dataset is employed to train deep learning models to support automated monitoring and smart decision-making in precision agriculture. This data set also has immense potential for real-time and practical use. It can be utilized to develop applications like mobile apps or automated systems where farmers can easily identify diseases at early stages and take immediate action, without the requirement of expert on-site knowledge. This data set can also be utilized with smart farming equipment like drones and sensors to track big fields in real time.
Why it matches plant phenotyping methodsカリフラワー葉の健康状態・病害状態を画像で分類するデータセット自体が研究の中心であり、植物病害表現型の取得・解析基盤に該当します。
abstractIn this publication, we are introducing a dataset containing a considerable number of images of cauliflower leaves.
Reproduction assets foundThe paper's core asset is its own cauliflower leaf disease image dataset (2,661 images, three classes), publicly deposited on Mendeley Data with DOI 10.17632/x995snz7p3.1 and a direct URL matching an allowed URL.Dataset · publiced from the following geographic locations:
1. Zailla, Singair, Manikganj
Latitude : 23°47′46.11"N Longitude : 90°13′15.73"E
2. Dattapara, Ashulia, Savar, Dhaka
Latitude : 23°52′26.3"N Longitude : 90°19′06.3"E
Data accessibility
Repository name: Mendeley Data
Data identification number: 10.17632/x995snz7p3.1
Direct URL to data: https://data.mendeley.com/datasets/x995snz7p3/1
The dataset is publicly available and can be accessed via the provided Mendeley Data repository link.
Related research article
None
1.
Value of the Data
•
This dataset holds high-resolution images of diseased cauliflower leaves infected with multiple diseases, which provide a wealth of material for the development and vaOpen asset ↗Mendeley Data · 10.17632/x995snz7p3.1lines:35-107Code / dataset availability confirmedEurope PMC · checked 6 Sept 2026
Verticillium wilt greatly hampers Chinese cabbage growth, causing significant yield limitations. Rapid and accurate detection of Verticillium wilt in the Chinese cabbage (Brassica rapa L. ssp. pekinensis) can provide significant agronomic benefits. Here, we propose a detection model, DSConv-GAN, which is based on images acquired by an unmanned aerial vehicle (UAV). Based on YOLOv8, with the addition of the dynamic snake convolution (DSConv) module and the improved loss function maximum possible distance intersection-over-union (MPDIoU), we acquired enhanced complex structures and global characteristics in Chinese cabbage images under different growth conditions. To reduce the difficulty of acquiring diseased Chinese cabbage data, a cycle-consistent generative adversarial network (CycleGAN) was used to simulate and generate images of the Verticillium wilt characteristics for multiple fields. The detection of lightly infected plants achieved precision, recall, mean average precision (mAP), and F1-score of 81.3, 86.6, 87.7, and 83.9%, respectively. DSConv-GAN outperforms other models in terms of precision, detection speed, robustness, and generalization. The model is combined with software to improve the practicability of the proposed method. Our results demonstrate DSConv-GAN to be an effective intelligent farming tool that provides early, rapid, and accurate detection of Chinese cabbage Verticillium wilt in complex growing environments.
Why it matches plant phenotyping methodsUAV画像から中国白菜の萎黄病状態を推定する検出モデルを開発・評価しており、植物病徴の取得・抽出手法が研究の中心である。
abstractHere, we propose a detection model, DSConv-GAN, which is based on images acquired by an unmanned aerial vehicle (UAV).
Reproduction assets foundThe paper's authors publicly release their analysis/detection code (DSConv-GAN model and monitoring software) via a GitHub repository, while the underlying UAV image dataset is only available upon request.Code · publicData will be made available upon request. The code could be downloaded from https://github.com/919449869coder/disease-detection-DSConv.git .Open asset ↗919449869coder/disease-detection-DSConvlines:151-175Code / dataset availability confirmedOpenAlex · checked 15 Sept 2026
Studies on the phenotypic traits and their associations in Chinese cabbage lack precise and objective digital evaluation metrics. Traditional assessment methods often rely on subjective evaluations and experience, compromising accuracy and reliability. This study develops an innovative, comprehensive trait evaluation method based on 3D point cloud technology, with the aim of enhancing the precision, reliability, and standardization of the comprehensive phenotypic traits of Chinese cabbage. By using multi-view image sequences and structure-from-motion algorithms, 3D point clouds of 50 plants from each of the 17 Chinese cabbage varieties were reconstructed. Color-based region growing and 3D convex hull techniques were employed to measure 30 agronomic traits. Comparisons between 3D point cloud-based measurements of the plant spread, plant height, leaf area, and leaf ball volume and traditional methods yielded R2 values greater than 0.97, with root mean square errors of 1.27 cm, 1.16 cm, 839.77 cm3, and 59.15 cm2, respectively. Based on the plant spread and plant height, a linear regression prediction of Chinese cabbage weights was conducted, yielding an R2 value of 0.76. Integrated optimization algorithms were used to test the parameters, reducing the measurement time from 55 min when using traditional methods to 3.2 min. Furthermore, in-depth analyses including variation, correlation, principal component analysis, and clustering analyses were conducted. Variation analysis revealed significant trait variability, with correlation analysis indicating 21 pairs of traits with highly significant positive correlations and 2 pairs with highly significant negative correlations. The top six principal components accounted for 90% of the total variance. Using the elbow method, k-means clustering determined that the optimal number of clusters was four, thus classifying the 17 cabbage varieties into four distinct groups. This study provides new theoretical and methodological insights for exploring phenotypic trait associations in Chinese cabbage and facilitates the breeding and identification of high-quality varieties. Compared with traditional methods, this system provides significant advantages in terms of accuracy, speed, and comprehensiveness, with its low cost and ease of use making it an ideal replacement for manual methods, being particularly suited for large-scale monitoring and high-throughput phenotyping.
Why it matches plant phenotyping methods中国白菜の表現型を3D点群から抽出する測定法を開発し、従来法との精度比較・検証および高速化を行っており、植物表現型測定が研究の中心である。
abstractThis study develops an innovative, comprehensive trait evaluation method based on 3D point cloud technology
Reproduction assets foundThe paper's phenotyping analysis code is explicitly deposited on a public GitHub repository with an authors' URL. The phenotype/trait measurement data themselves are only available upon request, so they do not qualify as a public asset.Code · publicapproach significantly streamlines the process, saving time and
enhancing efficiency by automating tasks which previously required extensive manual ef-
fort, thereby ensuring a more systematic and reliable method of phenotypic information
detection. The code used in this study can be accessed at the following GitHub repository:
https://github.com/chongchong123123/code (accessed on 18 October 2024).
2.4. Accuracy Analysis of Agronomic Parameter Measurements
In the course of agronomic trait measurement research, we utilized point cloud tech-
nology to measure key agronomic traits, including the plant height, plant spread, various
leaf dimensions (leaf length and leaf width), the width and thicOpen asset ↗chongchong123123/codepdf-raw-page:8 lines:1-62Code / dataset availability confirmedEurope PMC · checked 7 Sept 2026
Abstract In the realm of global food security, plants serve as the primary source of sustenance. However, plant diseases pose a significant threat to this security. The process of diagnosing these diseases forms the bedrock of disease control efforts. The precision and expediency of these diagnoses wield substantial influence over disease management and the consequent reduction of economic losses. Conversely, incorrect diagnoses can render interventions ineffective, leading to agricultural crop deterioration and compounding economic hardships for both farmers and their respective nations. This research endeavors to diagnose the prevalent crops in Jordan, as identified by the Jordanian Department of Statistics for the year 2019. These crops encompass four key agricultural varieties: cucumbers, tomatoes, lettuce, and cabbage. To facilitate this, a novel dataset known as "Jordan 22" was meticulously curated. Jordan 22 was painstakingly compiled through the collection of images featuring both diseased and healthy plants, captured within the confines of Jordanian farms. These images underwent meticulous classification by a panel of three agricultural specialists, well-versed in plant disease identification and prevention. The Jordan 22 dataset comprises a substantial size, amounting to 3210 images. Following the compilation of this dataset, a series of preprocessing steps were executed. These encompassed the standardization of image backgrounds and the uniformization of image dimensions. Furthermore, image augmentation techniques were applied to the dataset to expand its diversity. Subsequently, a deep learning model, the Convolutional Neural Network (CNN), was meticulously trained on the augmented dataset. The results yielded by the CNN were nothing short of remarkable, with a test accuracy rate reaching an impressive 0.9712. Optimal performance was observed when images were resized to 256x256 dimensions, and max pooling was employed in lieu of average pooling within the pooling layer. Furthermore, the initial convolutional layer was set at a size of 32, with subsequent convolutional layers standardized at 128 in size. In conclusion, this research represents a pivotal step towards enhancing plant disease diagnosis and, by extension, global food security. Through the creation of the Jordan 22 dataset and the meticulous training of a CNN model, we have achieved substantial accuracy in disease detection, paving the way for more effective disease management strategies in agriculture.
Why it matches plant phenotyping methods植物画像から病害状態を推定するCNNとデータセットを開発・評価しており、植物フェノタイピング手法が研究の中心である。
abstractThis research endeavors to diagnose the prevalent crops in Jordan
Reproduction assets foundThe paper's Jordan22 plant disease image dataset (2310 RGB leaf images of cucumber, tomato, cabbage, and lettuce collected in Jordan and expert-classified) is explicitly stated as openly available on the authors' public GitHub repository. No separate analysis code or trained model checkpoint is explicitly deposited.Dataset · publicThe data that support the findings of this study are openly available in [Jordan22_Dataset] at
[https://github.com/shahd1995913/Jordan22_Dataset], reference number [17].Open asset ↗Jordan22_Datasetpdf-page:25 lines:1-40Code / dataset availability confirmedCrossref · Europe PMC · checked 13 Sept 2026
Bangladesh's agricultural landscape is significantly influenced by vegetable cultivation, which substantially enhances nutrition, the economy, and food security in the nation. Millions of people rely on vegetable production for their daily sustenance, generating considerable income for numerous farmers. However, leaf diseases frequently compromise the yield and quality of vegetable crops. Plant diseases are a common impediment to global agricultural productivity, adversely affecting crop quality and yield, leading to substantial economic losses for farmers. Early detection of plant leaf diseases is crucial for improving cultivation and vegetable production. Common diseases such as Bacterial Spot, Mosaic Virus, and Downy Mildew often reduce vegetable plant cultivation and severely impact vegetable production and the food economy. Consequently, many farmers in Bangladesh struggle to identify the specific diseases, incurring significant losses. This dataset contains 12,643 images of widely grown crops in Bangladesh, facilitating the identification of unhealthy leaves compared to healthy ones. The dataset includes images of vegetable leaves such as Bitter Gourd (2223 images), Bottle Gourd (1803 images), Eggplants (2944 images), Cauliflowers (1598 images), Cucumbers (1626 images), and Tomatoes (2449 images). Each vegetable class encompasses several common diseases that affect cultivation. By identifying early leaf diseases, this dataset will be invaluable for farmers and agricultural researchers alike.
Why it matches plant phenotyping methods植物葉の画像から健康状態と病徴を識別するデータセットを提供しており、植物の病害状態を観測する再利用可能な画像ベースのフェノタイピング資源が中心です。
abstractThis dataset contains 12,643 images of widely grown crops in Bangladesh, facilitating the identification of unhealthy leaves compared to healthy ones.
Reproduction assets foundThe paper is a Data in Brief article describing a public smartphone image dataset of vegetable leaf diseases hosted on Mendeley Data, with an explicit direct URL matching an allowed URL.Dataset · publicRepository name: Mendeley Data
Data identification number: DOI: 10.17632/n67gctmjyj.3
Direct URL to data: https://data.mendeley.com/datasets/n67gctmjyj/3Open asset ↗Mendeley Data · 10.17632/n67gctmjyj.3lines:1-48Code / dataset availability confirmedOpenAlex · Europe PMC · checked 7 Sept 2026
Cauliflower cultivation is subject to high-quality control criteria during sales, which underlines the importance of accurate harvest timing. Using time series data for plant phenotyping can provide insights into the dynamic development of cauliflower and allow more accurate predictions of when the crop is ready for harvest than single-time observations. However, data acquisition on a daily or weekly basis is resource-intensive, making selection of acquisition days highly important. We investigate which data acquisition days and development stages positively affect the model accuracy to get insights into prediction-relevant observation days and aid future data acquisition planning. We analyze harvest-readiness using the cauliflower image time series of the GrowliFlower dataset. We use an adjusted ResNet18 classification model, including positional encoding of the data acquisition dates to add implicit information about development. The explainable machine learning approach GroupSHAP analyzes time points' contributions. Time points with the lowest mean absolute contribution are excluded from the time series to determine their effect on model accuracy. Using image time series rather than single time points, we achieve an increase in accuracy of 4%. GroupSHAP allows the selection of time points that positively affect the model accuracy. By using seven selected time points instead of all 11 ones, the accuracy improves by an additional 4%, resulting in an overall accuracy of 89.3%. The selection of time points may therefore lead to a reduction in data collection in the future.
Why it matches plant phenotyping methodsカリフラワー画像時系列を用いた収穫適期という植物状態の推定手法を開発・評価し、データ取得時点の選択とモデル精度を検証しているため、フェノタイピング手法が中心である。
abstractUsing time series data for plant phenotyping can provide insights into the dynamic development of cauliflower and allow more accurate predictions of when the crop is ready for harvest than single-time observations.
Reproduction assets foundThe paper analyzes the GrowliFlower cauliflower UAV image time series dataset and provides a public data availability link to the dataset metadata on phenoroam.phenorob.de. No author analysis code or trained model checkpoint is explicitly deposited.Dataset · publicPublicly available datasets were analyzed in this study. This data can be found at: https://phenoroam.phenorob.de/geonetwork/srv/eng/catalog.search#/metadata/cb328232-31f5-4b84-a929-8e1ee551d66a .Open asset ↗phenoroam.phenorob.de · cb328232-31f5-4b84-a929-8e1ee551d66alines:386-397Code / dataset availability confirmedCrossref · checked 15 Sept 2026
Growth monitoring of crops is a crucial aspect of precision agriculture, essential for optimal yield prediction and resource allocation. Traditional crop growth monitoring methods are labor-intensive and prone to errors. This study introduces an automated segmentation pipeline utilizing multi-date aerial images and ortho-mosaics to monitor the growth of cauliflower crops (Brassica Oleracea var. Botrytis) using an object-based image analysis approach. The methodology employs YOLOv8, a Grounding Detection Transformer with Improved Denoising Anchor Boxes (DINO), and the Segment Anything Model (SAM) for automatic annotation and segmentation. The YOLOv8 model was trained using aerial image datasets, which then facilitated the training of the Grounded Segment Anything Model framework. This approach generated automatic annotations and segmentation masks, classifying crop rows for temporal monitoring and growth estimation. The study’s findings utilized a multi-modal monitoring approach to highlight the efficiency of this automated system in providing accurate crop growth analysis, promoting informed decision-making in crop management and sustainable agricultural practices. The results indicate consistent and comparable growth patterns between aerial images and ortho-mosaics, with significant periods of rapid expansion and minor fluctuations over time. The results also indicated a correlation between the time and method of observation which paves a future possibility of integration of such techniques aimed at increasing the accuracy in crop growth monitoring based on automatically derived temporal crop row segmentation masks.
Why it matches plant phenotyping methods航空画像・オルソモザイクから作物列を自動セグメンテーションし、時系列の生育・成長を推定する画像解析パイプラインが研究の中心であり、植物形質の取得手法として適格。
abstractThis study introduces an automated segmentation pipeline utilizing multi-date aerial images and ortho-mosaics to monitor the growth of cauliflower crops
Reproduction assets foundThe paper's Data Availability Statement points to the authors' public Mendeley Data repository (GobhiSet, DOI 10.17632/dcjjcwc5dh.4), which contains the raw, manually, and automatically annotated RGB aerial images and ortho-mosaics of cauliflower used for the YOLOv8x-seg and Grounded SAM training and growth analysis inDataset · publicon of the manuscript.
Funding: This research received no external funding.
Data Availability Statement: No new data was created. However, the data that were used to perform
this research can be found in the article published at https://doi.org/10.1016/j.dib.2024.110506 and
available in the repository DOI: 10.17632/dcjjcwc5dh.4 (https://data.mendeley.com/drafts/dcjjcwc5dh).Conflicts of Interest: The authors declare no conflicts of interest.
References
1. Di, L.; Ustundag, B. Crop Growth Modeling and Yield Forecasting. In Agro-Geoinformatics; Springer: Cham, Switzerland, 2021.
[CrossRef]
2. Mithen, S.; Jenkins, E.; Jamjoum, K.; Nuimat, S.; Nortcliff, S.; Finlayson, B. Experimental crop growingOpen asset ↗data.mendeley.com · 10.17632/dcjjcwc5dh.4pdf-raw-page:17 lines:1-52Code / dataset availability confirmedOpenAlex · Crossref · Europe PMC · checked 7 Sept 2026
BACKGROUND: Image-based crop growth modeling can substantially contribute to precision agriculture by revealing spatial crop development over time, which allows an early and location-specific estimation of relevant future plant traits, such as leaf area or biomass. A prerequisite for realistic and sharp crop image generation is the integration of multiple growth-influencing conditions in a model, such as an image of an initial growth stage, the associated growth time, and further information about the field treatment. While image-based models provide more flexibility for crop growth modeling than process-based models, there is still a significant research gap in the comprehensive integration of various growth-influencing conditions. Further exploration and investigation are needed to address this gap. METHODS: We present a two-stage framework consisting first of an image generation model and second of a growth estimation model, independently trained. The image generation model is a conditional Wasserstein generative adversarial network (CWGAN). In the generator of this model, conditional batch normalization (CBN) is used to integrate conditions of different types along with the input image. This allows the model to generate time-varying artificial images dependent on multiple influencing factors. These images are used by the second part of the framework for plant phenotyping by deriving plant-specific traits and comparing them with those of non-artificial (real) reference images. In addition, image quality is evaluated using multi-scale structural similarity (MS-SSIM), learned perceptual image patch similarity (LPIPS), and Fréchet inception distance (FID). During inference, the framework allows image generation for any combination of conditions used in training; we call this generation data-driven crop growth simulation. RESULTS: Experiments are performed on three datasets of different complexity. These datasets include the laboratory plant Arabidopsis thaliana (Arabidopsis) and crops grown under real field conditions, namely cauliflower (GrowliFlower) and crop mixtures consisting of faba bean and spring wheat (MixedCrop). In all cases, the framework allows realistic, sharp image generations with a slight loss of quality from short-term to long-term predictions. For MixedCrop grown under varying treatments (different cultivars, sowing densities), the results show that adding these treatment information increases the generation quality and phenotyping accuracy measured by the estimated biomass. Simulations of varying growth-influencing conditions performed with the trained framework provide valuable insights into how such factors relate to crop appearances, which is particularly useful in complex, less explored crop mixture systems. Further results show that adding process-based simulated biomass as a condition increases the accuracy of the derived phenotypic traits from the predicted images. This demonstrates the potential of our framework to serve as an interface between a data-driven and a process-based crop growth model. CONCLUSION: The realistic generation and simulation of future plant appearances is adequately feasible by multi-conditional CWGAN. The presented framework complements process-based models and overcomes their limitations, such as the reliance on assumptions and the low exact field-localization specificity, by realistic visualizations of the spatial crop development that directly lead to a high explainability of the model predictions.
Why it matches plant phenotyping methods植物画像を生成し、そこから植物個体別形質を推定する二段階の画像ベース表現型解析フレームワークを開発・評価しており、表現型取得・推定手法が研究の中心である。
abstractWe present a two-stage framework consisting first of an image generation model and second of a growth estimation model, independently trained.
Reproduction assets foundThe paper's authors explicitly state that source code and links to the phenotyping datasets (Arabidopsis, GrowliFlower, MixedCrop) are publicly available in their GitHub repository, which implements the multi-conditional CWGAN crop growth simulation and growth estimation framework.Code · publicSource code and links to the datasets are publicly available at https://github.com/luked12/crop-growth-cgan .Open asset ↗luked12/crop-growth-cganlines:216-253Code / dataset availability confirmedEurope PMC · checked 14 Sept 2026
This research introduces an extensive dataset of unprocessed aerial RGB images and orthomosaics of Brassica oleracea crops, captured via a DJI Phantom 4. The dataset, publicly accessible, comprises 244 raw RGB images, acquired over six distinct dates in October and November of 2020 as well as 6 orthomosaics from an experimental farm located in Portici, Italy. The images, uniformly distributed across crop spaces, have undergone both manual and automatic annotations, to facilitate the detection, segmentation, and growth modelling of crops. Manual annotations were performed using bounding boxes via the Visual Geometry Group Image Annotator (VIA) and exported in the Common Objects in Context (COCO) segmentation format. The automated annotations were generated using a framework of Grounding DINO + Segment Anything Model (SAM) facilitated by YOLOv8x-seg pretrained weights obtained after training manually annotated images dated 8 October, 21 October, and 29 October 2020. The automated annotations were archived in Pascal Visual Object Classes (PASCAL VOC) format. Seven classes, designated as Row 1 through Row 7, have been identified for crop labelling. Additional attributes such as individual crop ID and the repetitiveness of individual crop specimens are delineated in the Comma Separated Values (CSV) version of the manual annotation. This dataset not only furnishes annotation information but also assists in the refinement of various machine learning models, thereby contributing significantly to the field of smart agriculture. The transparency and reproducibility of the processes are ensured by making the utilized codes accessible. This research marks a significant stride in leveraging technology for vision-based crop growth monitoring.
Why it matches plant phenotyping methods作物の生育モニタリングを目的としたRGB画像・オルソモザイクの公開データセットで、手動／自動アノテーションと成長モデリングを中心的に扱っているため、植物フェノタイピング手法・データセットに該当する。
abstractThis research introduces an extensive dataset of unprocessed aerial RGB images and orthomosaics of Brassica oleracea crops
Reproduction assets foundThe paper's own GobhiSet dataset (raw RGB images, orthomosaics, manual/automatic annotations, binary masks) and Python analysis scripts are publicly deposited on Mendeley Data with an explicit direct URL and DOI.Dataset · public0137
Longitude: 14; 20; 47.7701
Data post-processing and storage location: Department of Engineering, University of Campania ‘Luigi Vanvitelli,’ Aversa, Italy
Coordinates: 40.96846317808221, 14.208207168044456
Data accessibility
Repository name: Mendeley Data
Data identification number: 10.17632/dcjjcwc5dh.3
Direct URL to data: https://data.mendeley.com/datasets/dcjjcwc5dh/3
1.
Value of the Data
•
This dataset is a collection of multi-date aerial imagery of the Brassica oleracea var. Botrytis crop [ 1 ]. The images were acquired between the first and seventh weeks after sowing the cauliflower, with the intention of observing its growth over this period. The images were annotated with two typOpen asset ↗Mendeley Data · 10.17632/dcjjcwc5dh.3lines:50-75Code / dataset availability confirmedEurope PMC · checked 14 Sept 2026
In the dataset presented in this article, samples belonging to one of the following crops, apple, broccoli, leek, and mushroom, were measured by hyperspectral cameras in the visible/near-infrared spectral domain (430-900 nm). The dataset was compiled by putting together measurements from different calibrated hyperspectral imaging cameras and crops to facilitate the training of artificial intelligence models, helping to overcome the generalization problem of hyperspectral models. In particular, this dataset focuses on estimating dry matter content across various crops by a single model in a non-destructive way using hyperspectral measurements. This dataset contains extracted mean reflectance spectra for each sample (n=1028) and their respective dry matter content (%).
Why it matches plant phenotyping methods複数作物の果実・器官について、ハイパースペクトル画像から乾物含量を非破壊推定するデータセットを構築しており、形質取得・推定手法と再利用可能なベンチマークが研究の中心である。
abstractThe dataset was compiled by putting together measurements from different calibrated hyperspectral imaging cameras and crops to facilitate the training of artificial intelligence models, helping to overcome the generalization problem of hyperspectral models.
Reproduction assets foundThe paper is a data descriptor for the SpectroFood hyperspectral dataset; all five Zenodo deposits (meta-dataset plus per-crop hyperspectral image data) are public, paper-specific phenotype/trait datasets with direct URLs in the Specifications Table.Dataset · publicce), Rc: corrected hyperspectral image.
Data source location
Data are stored at Agricultural University of Athens (AUA) premises. Iera Odos 75, 11855 Athens, Greece, Department of Horticultural Engineering
Data accessibility
Repository name:Zenodo
Table data
Data identification number: 10.5281/zenodo.8362947
Direct URL to data: https://zenodo.org/record/8362947
Hyperspectral image data
1) Data identification number: 10.5281/zenodo.10301753
Direct URL to data: https://zenodo.org/records/10301753
2) Data identification number: 10.5281/zenodo.10302438
Direct URL to data: https://zenodo.org/records/10302438
3) Data identification number: 10.5281/zenodo.10302426
Direct URL to data: https:/Open asset ↗Zenodo · 10.5281/zenodo.8362947lines:1-65Dataset · publicOdos 75, 11855 Athens, Greece, Department of Horticultural Engineering
Data accessibility
Repository name:Zenodo
Table data
Data identification number: 10.5281/zenodo.8362947
Direct URL to data: https://zenodo.org/record/8362947
Hyperspectral image data
1) Data identification number: 10.5281/zenodo.10301753
Direct URL to data: https://zenodo.org/records/10301753
2) Data identification number: 10.5281/zenodo.10302438
Direct URL to data: https://zenodo.org/records/10302438
3) Data identification number: 10.5281/zenodo.10302426
Direct URL to data: https://zenodo.org/records/10302426
4) Data identification number: 10.5281/zenodo.10302386
Direct URL to data: https://zenodo.org/records/10302Open asset ↗Zenodo · 10.5281/zenodo.10301753lines:1-65Dataset · publicdo
Table data
Data identification number: 10.5281/zenodo.8362947
Direct URL to data: https://zenodo.org/record/8362947
Hyperspectral image data
1) Data identification number: 10.5281/zenodo.10301753
Direct URL to data: https://zenodo.org/records/10301753
2) Data identification number: 10.5281/zenodo.10302438
Direct URL to data: https://zenodo.org/records/10302438
3) Data identification number: 10.5281/zenodo.10302426
Direct URL to data: https://zenodo.org/records/10302426
4) Data identification number: 10.5281/zenodo.10302386
Direct URL to data: https://zenodo.org/records/10302386
1.
Value of the Data
•
Spectra were acquired using calibrated hyperspectral imaging systems under the sameOpen asset ↗Zenodo · 10.5281/zenodo.10302438lines:1-65Dataset · public8362947
Hyperspectral image data
1) Data identification number: 10.5281/zenodo.10301753
Direct URL to data: https://zenodo.org/records/10301753
2) Data identification number: 10.5281/zenodo.10302438
Direct URL to data: https://zenodo.org/records/10302438
3) Data identification number: 10.5281/zenodo.10302426
Direct URL to data: https://zenodo.org/records/10302426
4) Data identification number: 10.5281/zenodo.10302386
Direct URL to data: https://zenodo.org/records/10302386
1.
Value of the Data
•
Spectra were acquired using calibrated hyperspectral imaging systems under the same controlled conditions for four crops with high variation of dry matter values amongst the same crop and acrossOpen asset ↗Zenodo · 10.5281/zenodo.10302426lines:1-65Dataset · publicps://zenodo.org/records/10301753
2) Data identification number: 10.5281/zenodo.10302438
Direct URL to data: https://zenodo.org/records/10302438
3) Data identification number: 10.5281/zenodo.10302426
Direct URL to data: https://zenodo.org/records/10302426
4) Data identification number: 10.5281/zenodo.10302386
Direct URL to data: https://zenodo.org/records/10302386
1.
Value of the Data
•
Spectra were acquired using calibrated hyperspectral imaging systems under the same controlled conditions for four crops with high variation of dry matter values amongst the same crop and across all four.
•
The dry matter content of the four crops is the common variable when considering the quality of theOpen asset ↗Zenodo · 10.5281/zenodo.10302386lines:1-65Code / dataset availability confirmedEurope PMC · checked 7 Sept 2026
The use of unmanned aerial vehicles (UAVs) has facilitated crop canopy monitoring, enabling yield prediction by integrating regression models. However, the application of UAV-based data to individual-level harvest weight prediction is limited by the effectiveness of obtaining individual features. In this study, we propose a method that automatically detects and extracts multitemporal individual plant features derived from UAV-based data to predict harvest weight. We acquired data from an experimental field sown with 1196 Chinese cabbage plants, using two cameras (RGB and multi-spectral) mounted on UAVs. First, we used three RGB orthomosaic images and an object detection algorithm to detect more than 95% of the individual plants. Next, we used feature selection methods and five different multi-temporal resolutions to predict individual plant weights, achieving a coefficient of determination (R 2 ) of 0.86 and a root mean square error (RMSE) of 436 g/plant. Furthermore, we achieved predictions with an R 2 greater than 0.72 and an RMSE less than 560 g/plant up to 53 days prior to harvest. These results demonstrate the feasibility of accurately predicting individual Chinese cabbage harvest weight using UAV-based data and the efficacy of utilizing multi-temporal features to predict plant weight more than one month prior to harvest.
Why it matches plant phenotyping methodsUAV画像から個体特徴を自動抽出し、収穫重量という植物形質を予測する手法が研究の中心であるため。
abstractwe propose a method that automatically detects and extracts multitemporal individual plant features derived from UAV-based data to predict harvest weight.
Reproduction assets foundThe paper's analysis code is publicly available on GitHub with explicit availability language. The UAV imagery (RGB/multispectral orthomosaics and point cloud data) is only available upon reasonable request from the corresponding author, so it does not qualify as a public asset.Code · publicAll code associated with the current study is available at: https://github.com/anaguilarar/CC_Weight_Prediction .Open asset ↗anaguilarar/CC_Weight_Predictionlines:155-233Code / dataset availability confirmedEurope PMC · checked 14 Sept 2026
On-farm food loss (i.e., grade-out vegetables) is a difficult challenge in sustainable agricultural systems. The simplest method to reduce the number of grade-out vegetables is to monitor and predict the size of all individuals in the vegetable field and determine the optimal harvest date with the smallest grade-out number and highest profit, which is not cost-effective by conventional methods. Here, we developed a full pipeline to accurately estimate and predict every broccoli head size ( n > 3,000) automatically and nondestructively using drone remote sensing and image analysis. The individual sizes were fed to the temperature-based growth model and predicted the optimal harvesting date. Two years of field experiments revealed that our pipeline successfully estimated and predicted the head size of all broccolis with high accuracy. We also found that a deviation of only 1 to 2 days from the optimal date can considerably increase grade-out and reduce farmer's profits. This is an unequivocal demonstration of the utility of these approaches to economic crop optimization and minimization of food losses.
Why it matches plant phenotyping methodsドローンリモートセンシングと画像解析により、個々のブロッコリー頭部サイズを自動・非破壊推定するパイプラインを開発・検証しており、植物形質取得が中心的です。
abstractwe developed a full pipeline to accurately estimate and predict every broccoli head size ( n > 3,000) automatically and nondestructively using drone remote sensing and image analysis.
Reproduction assets foundThe authors' full phenotyping/analysis pipeline source code is publicly available on GitHub (UAVbroccoli). Original drone image data (224 GB for 2020, 72 GB for 2021) exist but are only available upon request via Google Drive. Generic tools (YOLOv5, BiSeNet, labelme, EasyIDP, scikit-image) are third-party libraries, soCode · publicurvey powered by ML/DL for sustainable agricultural development, there are some limitations to its use. First, our system is neither fully automated nor app-based; therefore, farmers without computer science backgrounds cannot use this system directly in their own fields. However, because the source code is open to the public ( https://github.com/UTokyo-FieldPhenomics-Lab/UAVbroccoli ), local agricultural institutes and agricultural companies are able to modify and use the system according to their target. This study is definitely not a one-stop solution, but is a pioneer in real agriculture applications. Second, unlike traditional manual methods with limited throughput, the proposed method Open asset ↗UTokyo-FieldPhenomics-Lab/UAVbroccolilines:291-292Code / dataset availability confirmedEurope PMC · checked 15 Sept 2026
Purple Chinese cabbage (PCC) has become a new breeding trend due to its attractive color and high nutritional quality since it contains abundant anthocyanidins. With the aim of rapid evaluation of PCC anthocyanidins contents and screening of breeding materials, a fast quantitative detection method for anthocyanidins in PCC was established using Near Infrared Spectroscopy (NIR). The PCC samples were scanned by NIR, and the spectral data combined with the chemometric results of anthocyanidins contents obtained by high-performance liquid chromatography were processed to establish the prediction models. The content of cyanidin varied from 93.5 mg/kg to 12,802.4 mg/kg in PCC, while the other anthocyanidins were much lower. The developed NIR prediction models on the basis of partial least square regression with the preprocessing of no-scattering mode and the first-order derivative showed the best prediction performance: for cyanidin, the external correlation coefficient (RSQ) and standard error of cross-validation (SECV) of the calibration set were 0.965 and 693.004, respectively; for total anthocyanidins, the RSQ and SECV of the calibration set were 0.966 and 685.994, respectively. The established models were effective, and this NIR method, with the advantages of timesaving and convenience, could be applied in purple vegetable breeding practice.
Why it matches plant phenotyping methods紫キャベツのアントシアニン含量という植物形質を、NIR分光とケモメトリクスで迅速推定する方法を開発・検証しており、表現型取得法が研究の中心である。
abstracta fast quantitative detection method for anthocyanidins in PCC was established using Near Infrared Spectroscopy (NIR).
Reproduction assets foundThe paper's supplementary material (Table S1) contains the paper-specific HPLC-measured anthocyanidin contents for the 106 purple Chinese cabbage samples used to build the NIR prediction models, and is publicly downloadable from MDPI. No author analysis code, spectral files, or trained model files are explicitly sharedSupplement · publicdual author(s) and contributor(s) and not of MDPI and/or the editor(s). MDPI and/or the editor(s) disclaim responsibility for any injury to people or property resulting from any ideas, methods, instructions or products referred to in the content.
Supplementary Materials
The following supporting information can be downloaded at: https://www.mdpi.com/article/10.3390/foods12091922/s1 , Table S1: Contents of anthocyanidins in purple Chinese cabbage analyzed by high-performance liquid chromatography (mg/kg).
Click here for additional data file.
Author Contributions
Conceptualization, D.-S.Z. and H.-J.H.; software, G.-M.L.; validation, Y.-Q.W. and L.-P.H.; formal analysis, G.-M.L. and X.-Z.Z.; invOpen asset ↗lines:56-122Code / dataset availability confirmedEurope PMC · OpenAlex · Crossref · checked 8 Sept 2026
Premise With modern advances in genetic sequencing technology, plant phenotyping has become a substantial bottleneck in crop improvement programs. Traditionally, researchers have manually measured phenotypic traits to help determine genotype-phenotype relationships, but manual measurements can be time consuming and expensive. Recently, automated phenotyping systems have increased the spatial and temporal density of measurements, but most of these systems are extremely expensive and require specialized expertise. In the present paper, we develop and validate a low-cost, scalable, high-throughput phenotyping (HTP) system for automating the measurement of foliar area and greenness. Methods During a greenhouse experiment on the effects of abiotic stress on Brassica rapa , we collected images of hundreds of plants every hour for over a month with a system that cost approximately US$1000. Results In comparison with manually acquired images, this HTP system was able to produce similar estimates of foliar area and greenness, developmental trends, and treatment effects. Foliar area was correlated between the two image sets, but greenness was not. Discussion These findings highlight the potential of HTP systems built from low-cost hardware and freely available software. Future work can use this system to investigate genotype-environment interactions and the genetic loci underlying morphological changes resulting from abiotic stress.
Why it matches plant phenotyping methods低コストの画像ベース高スループット表現型計測システムを開発・検証し、葉面積と緑色度を自動推定することが研究の中心であるため。
abstractwe develop and validate a low-cost, scalable, high-throughput phenotyping (HTP) system for automating the measurement of foliar area and greenness.
Reproduction assets foundThe paper's data availability statement deposits both the phenotyping data (images/measurements) and the analysis scripts on Zenodo with explicit public DOIs, making both paper-specific assets directly actionable.Dataset · publicAll of the data and analysis scripts have been deposited to Zenodo (data: https://doi.org/10.5281/zenodo.5725224 ; scripts: https://doi.org/10.5281/zenodo.6366716 ).Open asset ↗Zenodo · 10.5281/zenodo.5725224lines:188-243Code · publicAll of the data and analysis scripts have been deposited to Zenodo (data: https://doi.org/10.5281/zenodo.5725224 ; scripts: https://doi.org/10.5281/zenodo.6366716 ).Open asset ↗Zenodo · 10.5281/zenodo.6366716lines:188-243Code / dataset availability confirmedEurope PMC · checked 8 Sept 2026
Precise and site-specific nitrogen (N) fertilizer management of vegetables is essential to improve the N use efficiency considering temporal and spatial fertility variations among fields, while the current N fertilizer recommendation methods are proved to be time- and labor-consuming. To establish a site-specific N topdressing algorithm for bok choy ( Brassica rapa subsp. chinensis ), using a hand-held GreenSeeker canopy sensor, we conducted field experiments in the years 2014, 2017, and 2020. Two planting densities, viz, high (123,000 plants ha -1 ) in Year I and low (57,000 plants ha -1 ) in Year II, whereas, combined densities in Year III were used to evaluate the effect of five N application rates (0, 45, 109, 157, and 205 kg N ha -1 ). A robust relationship was observed between the sensor-based normalized difference vegetation index (NDVI), the ratio vegetation index (RVI), and the yield potential without topdressing (YP 0 ) at the rosette stage, and 81-84% of the variability at high density and 76-79% of that at low density could be explained. By combining the densities and years, the R 2 value increased to 0.90. Additionally, the rosette stage was identified as the earliest stage for reliably predicting the response index at harvest (RI Harvest ), based on the response index derived from NDVI (RI NDVI ) and RVI (RI RVI ), with R 2 values of 0.59-0.67 at high density and 0.53-0.65 at low density. When using the combined results, the RI RVI performed 6.12% better than the RI NDVI , and 52% of the variability could be explained. This study demonstrates the good potential of establishing a sensor-based N topdressing algorithm for bok choy, which could contribute to the sustainable development of vegetable production.
Why it matches plant phenotyping methods携帯型キャノピーセンサーのNDVI/RVIから収量ポテンシャルと施肥応答を推定するアルゴリズムを開発・検証しており、植物形質推定手法が研究の中心です。
abstractTo establish a site-specific N topdressing algorithm for bok choy ( Brassica rapa subsp. chinensis ), using a hand-held GreenSeeker canopy sensor, we conducted field experiments in the years 2014, 2017, and 2020.
Reproduction assets found本文中に内容が明示された植物フェノタイピング関連の補足表と、その公開リンクを確認しました。Supplement · publicSupplementary Table 2 ), the empirical exponential model was used to determine the relationship between YP 0 and the sensor-based vegetation indices (NDVI and RVI) for bok choy across growth stages ( Table 3 ).Open asset ↗lines:391-483Code / dataset availability confirmedEurope PMC · checked 14 Sept 2026
Cauliflower, a winter seasoned vegetable that originated in the Mediterranean region and arrived in Europe at the end of the 15th century, takes the lead in production among all vegetables. It's high in fiber and can keep us hydrated, and have medicinal properties like the chemical glucosinolates, which may help prevent cancer. If proper care is not given to the plants, several significant diseases can affect the plants, reducing production, quantity, and quality. Plant disease monitoring by hand is extremely tough because it demands a great deal of effort and time. Early detection of the diseases allows the agriculture sector to grow cauliflower more efficiently. In this scenario, an insightful and scientific dataset can be a lifesaver for researchers looking to analyze and observe different diseases in cauliflower development patterns. So, in this work, we present a well-organized and technically valuable dataset "VegNet' to effectively recognize conditions in cauliflower plants and fruits. Healthy and disease-affected cauliflower head and leaves by black rot,downy mildew, and bacterial spot rot are included in our suggested dataset. The images were taken manually from December 20th to January 15th, when the flowers were fully blown, and most of the diseases were observed clearly. It is a well-organized dataset to develop and validate machine learning-based automated cauliflower disease detection algorithms. The dataset is hosted by the Institute - National Institute of Textile Engineering and Research (NITER),the Department of Computer Science and Engineering and is available at the link following: https://data.mendeley.com/datasets/t5sssfgn2v/3.
Why it matches plant phenotyping methodsカリフラワーの病害状態を画像で識別するためのデータセットを構築・提供しており、植物の病害表現型取得が研究の中心である。
abstractwe present a well-organized and technically valuable dataset "VegNet' to effectively recognize conditions in cauliflower plants and fruits.
Reproduction assets foundThe paper is a data descriptor for VegNet, a public dataset of cauliflower disease images hosted on Mendeley Data with an explicit permanent identifier and direct link, directly reproducing the paper's plant image measurements.Dataset · publicrly. It is a well-organized dataset to develop and validate machine learning-based automated cauliflower disease detection algorithms. The dataset is hosted by the Institute – National Institute of Textile Engineering and Research (NITER),the Department of Computer Science and Engineering and is available at the link following: https://data.mendeley.com/datasets/t5sssfgn2v/3 .
Keywords: Computer vision system, Agriculture, Feature extraction, Machine learning
status released display-pdf yes is-olf no is-manuscript no is-preprint no is-journal-matter no is-scanned no is-retracted no
Received 2022 May 5; Revised 2022 Jun 8; Accepted 2022 Jun 21; Collection date 2022 Aug.
Specifications Table
SOpen asset ↗lines:1-47Code / dataset availability confirmedEurope PMC · checked 8 Sept 2026
Ultraviolet-B (UV-B, 280-315 nm) radiation has been known as an elicitor to enhance bioactive compound contents in plants. However, unpredictable yield is an obstacle to the application of UV-B radiation to controlled environments such as plant factories. A typical three-dimensional (3D) plant structure causes uneven UV-B exposure with leaf position and age-dependent sensitivity to UV-B radiation. The purpose of this study was to develop a model for predicting phenolic accumulation in kale ( Brassica oleracea L. var. acephala ) according to UV-B radiation interception and growth stage. The plants grown under a plant factory module were exposed to UV-B radiation from UV-B light-emitting diodes with a peak at 310 nm for 6 or 12 h at 23, 30, and 38 days after transplanting. The spatial distribution of UV-B radiation interception in the plants was quantified using ray-tracing simulation with a 3D-scanned plant model. Total phenolic content (TPC), total flavonoid content (TFC), total anthocyanin content (TAC), UV-B absorbing pigment content (UAPC), and the antioxidant capacity were significantly higher in UV-B-exposed leaves. Daily UV-B energy absorbed by leaves and developmental age was used to develop stepwise multiple linear regression models for the TPC, TFC, TAC, and UAPC at each growth stage. The newly developed models accurately predicted the TPC, TFC, TAC, and UAPC in individual leaves with R 2 > 0.78 and normalized root mean squared errors of approximately 30% in test data, across the three growth stages. The UV-B energy yields for TPC, TFC, and TAC were the highest in the intermediate leaves, while those for UAPC were the highest in young leaves at the last stage. To the best of our knowledge, this study proposed the first statistical models for estimating UV-B-induced phenolic contents in plant structure. These results provided the fundamental data and models required for the optimization process. This approach can save the experimental time and cost required to optimize the control of UV-B radiation.
Why it matches plant phenotyping methods3DスキャンとレイトレーシングによるUV-B吸収量の推定、およびフェノール含量予測モデルの開発が研究の中心であり、植物形質の計測・推定手法に該当する。
abstractThe spatial distribution of UV-B radiation interception in the plants was quantified using ray-tracing simulation with a 3D-scanned plant model.
Reproduction assets found本文中に内容が明示された植物フェノタイピング関連の補足表と、その公開リンクを確認しました。Supplement · publicSupplementary Table 1
Relative growth rate and expansion rate of leaf groups, and the assigned leaf order in kale plants at 23, 30, and 38 DAT.Open asset ↗lines:612-691Code / dataset availability confirmedEurope PMC · Crossref · checked 15 Sept 2026
Background Unmanned aerial vehicle (UAV)-based image retrieval in modern agriculture enables gathering large amounts of spatially referenced crop image data. In large-scale experiments, however, UAV images suffer from containing a multitudinous amount of crops in a complex canopy architecture. Especially for the observation of temporal effects, this complicates the recognition of individual plants over several images and the extraction of relevant information tremendously. Results In this work, we present a hands-on workflow for the automatized temporal and spatial identification and individualization of crop images from UAVs abbreviated as "cataloging" based on comprehensible computer vision methods. We evaluate the workflow on 2 real-world datasets. One dataset is recorded for observation of Cercospora leaf spot-a fungal disease-in sugar beet over an entire growing cycle. The other one deals with harvest prediction of cauliflower plants. The plant catalog is utilized for the extraction of single plant images seen over multiple time points. This gathers a large-scale spatiotemporal image dataset that in turn can be applied to train further machine learning models including various data layers. Conclusion The presented approach improves analysis and interpretation of UAV data in agriculture significantly. By validation with some reference data, our method shows an accuracy that is similar to more complex deep learning-based recognition techniques. Our workflow is able to automatize plant cataloging and training image extraction, especially for large datasets.
Why it matches plant phenotyping methodsUAV画像から個体を時空間的に同定・個別化し、植物画像データセットを抽出するコンピュータビジョン手法が研究の中心であり、精度検証も行っている。
abstractwe present a hands-on workflow for the automatized temporal and spatial identification and individualization of crop images from UAVs
Reproduction assets foundThe paper's authors publicly released their plant cataloging workflow code on GitHub and deposited a supporting subset of the sugar beet UAV image data with code snapshots in GigaDB (10.5524/102225). The GitHub repository URL is in the allowed list; the GigaDB DOI is not, so only the code asset is listed with an exact-Code · publicponding data. By automatizing the plant cataloging and providing a data framework, our work helps to exploit the full potential of UAV imaging in agricultural contexts.
Availability of Source Code
The source code of our workflow is available in the following repository:
Project name: Plant Cataloging Workflow
GitHub repository: https://github.com/mrcgndr/plant_cataloging_workflow
RRID: SCR_022276
Operating system(s): Platform independent (with conda), Linux (with Docker)
Programming language: Python (3.9 or higher)
License: Apache License 2.0
Data Availability
A subset of the sugar beet data is available in order to run the workflow and reproduce our results. The data have been uploaded to tOpen asset ↗https://github.com/mrcgndr/plant_cataloging_workflowlines:172-190Code / dataset availability confirmedOpenAlex · Europe PMC · checked 8 Sept 2026
BACKGROUND: Cabbage white butterflies (Pieris spp.) can be severe pests of Brassica crops such as Chinese cabbage, Pak choi (Brassica rapa) or cabbages (B. oleracea). Eggs of Pieris spp. can induce a hypersensitive response-like (HR-like) cell death which reduces egg survival in the wild black mustard (B. nigra). Unravelling the genetic basis of this egg-killing trait in Brassica crops could improve crop resistance to herbivory, reducing major crop losses and pesticides use. Here we investigated the genetic architecture of a HR-like cell death induced by P. brassicae eggs in B. rapa. RESULTS: A germplasm screening of 56 B. rapa accessions, representing the genetic and geographical diversity of a B. rapa core collection, showed phenotypic variation for cell death. An image-based phenotyping protocol was developed to accurately measure size of HR-like cell death and was then used to identify two accessions that consistently showed weak (R-o-18) or strong cell death response (L58). Screening of 160 RILs derived from these two accessions resulted in three novel QTLs for Pieris brassicae-induced cell death on chromosomes A02 (Pbc1), A03 (Pbc2), and A06 (Pbc3). The three QTLs Pbc1-3 contain cell surface receptors, intracellular receptors and other genes involved in plant immunity processes, such as ROS accumulation and cell death formation. Synteny analysis with A. thaliana suggested that Pbc1 and Pbc2 are novel QTLs associated with this trait, while Pbc3 also contains an ortholog of LecRK-I.1, a gene of A. thaliana previously associated with cell death induced by a P. brassicae egg extract. CONCLUSIONS: This study provides the first genomic regions associated with the Pieris egg-induced HR-like cell death in a Brassica crop species. It is a step closer towards unravelling the genetic basis of an egg-killing crop resistance trait, paving the way for breeders to further fine-map and validate candidate genes.
Why it matches plant phenotyping methodsHR様細胞死のサイズを定量する画像ベース表現型解析プロトコルを開発し、遺伝解析に実質的に使用しているため、植物フェノタイピング手法が中心的です。
abstractAn image-based phenotyping protocol was developed to accurately measure size of HR-like cell death
Reproduction assets foundThe authors state that datasets and scripts used for data analysis (including phenotypic data of the germplasm screening and RIL QTL experiments) are publicly available in a Zenodo repository, which is a paper-specific, publicly actionable asset.Dataset · publicDatasets and scripts used for data analysis are also available in a Zenodo repository ( https://doi.org/10.5281/zenodo.6014948 ).Open asset ↗Zenodo · 10.5281/zenodo.6014948lines:176-273Code / dataset availability confirmedEurope PMC · OpenAlex · checked 14 Sept 2026
The growth and the harvestability of a broccoli crop is monitored by the size of the broccoli head. This size estimation is currently done by humans, and this is inconsistent and expensive. The goal of our work was to develop a software algorithm that can estimate the size of field-grown broccoli heads based on RGB-Depth (RGB-D) images. For the algorithm to be successful, the problem of occlusion must be solved, which is the partial visibility of the broccoli head due to overlapping leaves. This partial visibility causes sizing errors. In this research, we studied the use of deep-learning algorithms to deal with occlusions. We specifically applied the Occlusion Region-based Convolutional Neural Network (ORCNN) that segmented both the visible and the amodal region of the broccoli head (which is the visible and the occluded region combined). We hypothesised that ORCNN, with its amodal segmentation, can improve the size estimation of occluded broccoli heads. The ORCNN sizing method was compared with a Mask R–CNN sizing method that only used the visible broccoli region to estimate the size. The sizing performance of both methods was evaluated on a test set of 487 broccoli images with systematic levels of leaf occlusion. With a mean sizing error of 6.4 mm, ORCNN outperformed Mask R–CNN, which had a mean sizing error of 10.7 mm. Furthermore, ORCNN had a significantly lower absolute sizing error on 161 heavily occluded broccoli heads with an occlusion rate between 50% and 90%. Our software and data set are available on https://git.wur.nl/blok012/sizecnn.
Why it matches plant phenotyping methodsRGB-D画像と深層学習によりブロッコリー頭部サイズを推定する手法を開発し、Mask R-CNNと比較検証しているため、植物形質取得法が研究の中心です。
abstractThe goal of our work was to develop a software algorithm that can estimate the size of field-grown broccoli heads based on RGB-Depth (RGB-D) images.
Brassica oleracea is an important crop species that at early growth stages may exhibit failure of the apical growing point, an abnormality called “blindness”. The occurrence of blindness is promoted by exposure to low temperatures during imbibition and germination, but the causes of sensitivity to such conditions are unknown. We combined three analytical seed technology instruments to explore seed physical properties that are highly correlated with quality parameters and might be used directly for grading or sorting seed lots into subpopulations varying in potential susceptibility to blindness. For image analysis, we used the VideometerLab instrument, which can scan 19 wavelengths from ultraviolet to infrared and utilize that information in any combination to potentially identify unique criteria related to seed quality. The iXeed CF Analyzer was utilized to obtain chlorophyll fluorescence values for individual seeds. Chlorophyll contents of many seeds can be used as an indicator of seed maturity, a major contributor to seed quality. Finally, oxygen consumption measurements of individual seeds as obtained with the Q2 instrument are highly correlated with their performance under a wide variety of conditions. Six Brassica seed lots differed in their susceptibility to induction of blindness or loss of viability due to 48 h hydrated incubation at 1.5 ∘C. Analysis of physical and respiratory parameters identified some measurements that were highly correlated with the occurrence of blindness. Higher chlorophyll content, as detected by the CF-Mobile and certain wavelengths in the Videometer, was associated with greater occurrence of blindness or death following the induction treatment, suggesting that more immature seeds may be susceptible to blindness. Further research is required, but methods to detect and sort such seeds based on physical characteristics appear to be feasible.
Why it matches plant phenotyping methods種子の画像・蛍光・呼吸測定を組み合わせ、物理特性から発芽品質やblindness感受性を評価・選別する方法が研究の中心であり、単なる生物学的結果測定ではない。
abstractWe combined three analytical seed technology instruments to explore seed physical properties that are highly correlated with quality parameters and might be used directly for grading or sorting seed lots into subpopulations varying in potential susceptibility to blindness.
Reproduction assets foundThe paper's individual-seed phenotyping measurements (chlorophyll fluorescence, multispectral imaging, Q2 respiration, plant blindness scores) are consolidated in Supplemental Table S1 (Seed parameters database) and related supplements, publicly hosted on the MDPI article site. No author analysis code was deposited; CRDataset · publicSupplementary Materials: The following are available at https://www.mdpi.com/2077-0472/11/3
/220/s1, Table S1: Seed parameters database, Table S2: Q2 parameters, Table S3: MFA EigenvaluesOpen asset ↗pdf-page:20 lines:1-58