High-throughput phenotyping of lettuce seedlings is highly prone to background confusion because the seedlings are small, have weak textural features, and exhibit spectral reflectance similar to that of the substrate. Traditional single-visual-modality approaches struggle to achieve reliable structural and physiological characterization simultaneously under the repetitive backgrounds and dense arrangements typical of greenhouse tray cultivation. To address these challenges, we establish a multimodal 3D phenotyping framework tailored for controlled agriculture environments, enabling the quantification of structural and physiological characteristics of lettuce seedlings. This framework is based on an unmanned ground vehicle (UGV) platform integrating a RGBD camera and a quad-band multispectral sensor which are rigidly coupled and synchronously triggered. An alignment module based on established feature matching algorithm is introduced to register the misalignment between source multispectral and RGBD images. Subsequently, we design a novel dual-backbone instance segmentation network, MS-SegNet, to enhance segmentation accuracy by hierarchically fusing geometric information with multispectral features. A robust 3D metric pose estimation pipeline, incorporating standard SfM initialization, scale recovery, and generalized ICP refinement, is constructed to generate 3D point clouds with spectral attributes and semantic labels. Finally, key structural and physiological phenotype parameters of each seedling are calculated based on the 3D semantic multispectral point clouds. Experiments demonstrate that MS-SegNet achieves significant advantages in instance segmentation of lettuce seedlings with mAP@50:95 = 0.854. The metric 3D pose estimation pipeline exhibits reliable performance under complex controlled conditions. The quality of the 3D reconstructions is indirectly validated through downstream structural trait extraction. The estimated seedling height and crown width show high correlation with manual measurements, achieving R 2 values of 0.8379 and 0.918, and RMSE values of 10.94 mm and 11.56 mm, respectively. Overall, by systematically integrating these adapted components with the novel segmentation architecture, this framework achieves stable performance improvements in 3D reconstruction, instance segmentation, and phenotypic analysis under greenhouse conditions. It provides a scalable, integrated technical solution for non-destructive, high-throughput phenotyping of crop seedlings in controlled environments.
Why it matches plant phenotyping methodsRGBD・マルチスペクトル・UGVを統合した3Dフェノタイピング基盤を開発し、分割・再構成・構造／生理形質抽出を検証しており、フェノタイピング手法が研究の中心である。
abstractwe establish a multimodal 3D phenotyping framework tailored for controlled agriculture environments, enabling the quantification of structural and physiological characteristics of lettuce seedlings.
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.
To improve the instance segmentation accuracy of mango fruits and peduncles in complex mountainous orchard scenes and provide visual decision support for robotic harvesting, this study proposes a model-driven perception and picking-point localization method. Specifically, an RGB-D mango dataset was constructed under natural orchard conditions, covering strong illumination, shadows, backlighting, fruit overlap, branch and leaf occlusion, and peduncle crossing. An improved lightweight instance segmentation model named SHS-YOLOv8n-seg was then developed based on YOLOv8n-seg. StarNet_s1 was introduced as the backbone to enhance feature extraction under complex backgrounds. Furthermore, a high-frequency and spatial perception feature pyramid network was adopted to strengthen multi-scale feature fusion and improve the representation of slender peduncles. The SPPF module was used to expand the receptive field, and the parameter-free SimAM attention mechanism was introduced to enhance target responses while suppressing background interference. In the single-run comparison, the proposed model achieved Precision, Recall, mAP@50, and mAP@50:95 values of 89.62%, 88.17%, 90.19%, and 67.94%, respectively. Compared with the baseline YOLOv8n-seg model, these values increased by 2.42, 2.90, 2.75, and 2.57 percentage points, respectively. Moreover, fruit–peduncle matching, geometric constraints, RGB-D depth information, and PCA-based local direction estimation were combined to infer the picking point and recover its 3D coordinates from the segmentation masks. In the evaluation of 57 RGB-D images, the picking point position accuracy reached 98.2%, while the local peduncle direction accuracy reached 91.2%. Overall, the proposed method can accurately segment mango fruits and peduncles in complex natural environments and convert the segmentation results into picking point positions, 3D coordinates, and local direction information, thereby providing theoretical and technical support for intelligent mango harvesting robots.
Why it matches plant phenotyping methodsマンゴー果実・果梗の画像セグメンテーションと3D形状情報の抽出手法を開発・評価しており、単なる収穫対象の位置検出を超えて、再利用可能な植物器官の形態情報を取得する方法が中心である。
abstractfruit–peduncle matching, geometric constraints, RGB-D depth information, and PCA-based local direction estimation were combined to infer the picking point and recover its 3D coordinates from the segmentation masks
The objective of this study was to develop a 3D plant modeling strategy that enables camera pose recovery from segmented plant images and the reconstruction of an initial point cloud. A lightweight, contour-aware framework leverages the view-consistent and surface-oriented representation of 2D Gaussian Splatting, making it suitable for plant surface reconstruction under the Plant-to-Camera mode. A contour-weighted Laplacian regularization suppresses depth discontinuities around plant boundaries, while simplified Gaussian primitives improve computational efficiency without compromising geometric fidelity. Organ-level semantics are integrated into the reconstructed geometry to distinguish plant organs such as leaves, stems, and ears. On maize and wheat datasets, our method outperformed existing approaches in terms of morphological fidelity, organ-level structural consistency, and processing speed, demonstrating its suitability for plant phenotyping
Why it matches plant phenotyping methods植物器官の3D再構成と形態情報抽出を目的とする計算手法を開発し、既存法と形態忠実度・器官構造整合性・処理速度で比較評価しており、フェノタイピング手法が中心である。
abstractThe objective of this study was to develop a 3D plant modeling strategy that enables camera pose recovery from segmented plant images and the reconstruction of an initial point cloud.
Plants are geometrically and topologically complex objects, and methods and devices that produce plant point clouds often miss parts due to self occlusions, making further analysis, such as phenotypic trait extraction or 3D reconstruction, difficult. We introduce A-Occ-Plant , a novel method for point cloud completion. The first novelty of our algorithm is converting point clouds into a set of images, which are then completed using 2D amodal segmentation. The images are then converted into a complete point cloud by using view-consistent Gaussian splats. The second novelty is the use of a coarse-to-fine hierarchical Transformer with cross-scale attention. The completed soft masks are fused into a continuous 3D density field using Gaussian splatting, removing the need for external pose estimation or fixed-size inputs. We introduce a synthetic dataset using a procedural model and a real-world plant reconstruction benchmark with artificially generated occlusions. We further benchmark A-Occ-Plant against representative 3D point-cloud completion methods, demonstrate that it recovers downstream phenotypic traits (leaf count, leaf angle, plant height), and show that it generalizes to another crops (soybean). A-Occ-Plant achieves a 264.8% improvement in LPIPS and an 8.3% gain in SSIM compared to the current state of the art, while using only 2.3% of the parameters and running 39.4× faster. We release our code at https://github.com/JaeLee18/PlantPhenomics_Occlusion.
Why it matches plant phenotyping methods植物の遮蔽点群を補完し、葉数・葉角度・草丈という表現型形質を復元する手法を開発しており、データセット作成とベンチマーク検証も中心的に行っている。
abstractWe introduce A-Occ-Plant , a novel method for point cloud completion.
Reproduction assets foundThe paper explicitly releases authors' code and sample data (inference code, sample data for reproducing results) via a Google Drive project download and a GitHub repository, both with explicit availability statements and public URLs.Code · publicThe full code and data at https://github.com/JaeLee18/PlantPhenomics_Occlusion .Open asset ↗JaeLee18/PlantPhenomics_Occlusionlines:386-410Plant phenotyping relevance match · UnverifiedCrossref · checked 13 Sept 2026
Abstract Accurate detection of strawberry fruit, reliable ripeness estimation, and precise localization of the picking point are essential for automated harvesting and yield prediction in smart farming. However, real-world environments introduce significant challenges, including occlusion, illumination variability, and high visual similarity between ripeness stages. Deep learning (DL)-based object detection methods have become the dominant approach to address these issues. This paper presents a systematic review of 60 peer-reviewed studies published between 2023 and 2026, focusing on detecting strawberries and their ripeness using the YOLO family (v5–v11) of DL algorithms. The studies are analyzed with respect to dataset characteristics, preprocessing and augmentation strategies, model architectures, and evaluation protocols. Our results show a clear dominance of YOLOv8, used in 28 (46.7%) of the 60 reviewed works, due to its real-time capability and architectural flexibility. Despite its short history, YOLOv11 has been adopted in 13 studies (21.7%) owing to its balanced precision and computational efficiency. Hybrid CNN–ViT models that integrate Transformer modules or networks into YOLO are gaining attention (8 studies, 13.3%) and show improved performance in complex scenarios, but they still incur higher computational cost. However, we identify critical methodological issues that affect the validity of reported results. In particular, the improper application of data augmentation prior to dataset splitting — a practice observed in precisely one-third of the reviewed studies — poses a significant risk of data leakage and can result in overly optimistic performance estimates. Additional challenges include inconsistent evaluation metrics, limited dataset diversity, and a lack of standardized benchmarks. This review provides a structured overview of current approaches, a critical assessment of existing research practices, and actionable guidance for developing robust, deployment-ready DL solutions for precision agriculture.
Why it matches plant phenotyping methodsイチゴ果実の検出・成熟度推定を対象とする画像ベース手法の系統的レビューであり、データセット、モデル、評価法、データリークやベンチマーク不足を批判的に検討しているため、フェノタイピング手法レビューとして中心的である。
abstractThis paper presents a systematic review of 60 peer-reviewed studies published between 2023 and 2026, focusing on detecting strawberries and their ripeness using the YOLO family (v5–v11) of DL algorithms.
Abstract Rice ( Oryza sativa L.) tiller angle is an important trait that influences plant architecture, canopy light interception, and yield potential. In this study, we proposed a deep learning‐based pipeline for automated measurement of rice tiller angle and plant base width using unmanned aerial vehicle (UAV) imagery. Our method leverages keypoint detection models to estimate tiller angles efficiently and accurately. We collected and annotated a dataset of UAV‐captured rice plant images for tiller angle estimation. We demonstrated that our approach provides scalable and precise measurement for keypoints under real‐world field conditions, achieving a mean average precision (mAP)@50 of 0.982 and a mAP@50:95 of 0.859 on the test set. Predicted tiller angle distribution aligns well with human annotations, with a mean absolute error of 5.3° across a range of 10.7°–27.8° and a Pearson's correlation coefficient of 0.64, offering acceptable accuracy for tiller angle measurement in real‐world agricultural settings. Additionally, the predicted plant base width ranges from 2.8 to 5.5 cm, with a mean absolute error of 1.02 cm compared to human annotations, highlighting the model's capability for precise spatial analysis. Significant differences in tiller angle and plant base width were detected among 27 rice genotypes. These results validate the proposed pipeline's potential for accurate and efficient differentiation of plant architecture traits. This research is the first to measure the rice tiller angle directly from UAV images. It lays a foundation for automated phenotyping of plant architecture traits and has the potential for integration into plant phenotyping frameworks to further promote artificial intelligence‐driven rice research and production.
Why it matches plant phenotyping methodsUAV画像と深層学習により、イネの分げつ角度・株元幅という植物形態形質を自動推定する手法を開発・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。
abstractwe proposed a deep learning‐based pipeline for automated measurement of rice tiller angle and plant base width using unmanned aerial vehicle (UAV) imagery.
Field / plotFlowerObject detectionPose / keypoint estimation
Agricultural engineering informatics is playing an increasingly important role in enabling intelligent perception, decision-making, and automated operations in modern horticultural production systems. Within this context, accurate visual perception of reproductive structures is essential for agricultural informatization tasks such as flowering-stage monitoring, precision pollination, and information-driven fruit-set management in chili cultivation. However, reliable detection and pose-aware recognition of chili flowers remain challenging because of small target size, dense distribution, foliage occlusion, and illumination variability in natural or semi-controlled environments. To address these challenges, this study proposes a lightweight and robust edge vision framework, termed CFPR-YOLO, for chili flower detection and pose-aware perception under complex agricultural conditions. Built upon an improved YOLOv11n architecture, the proposed framework incorporates EfficientFormerV2 to strengthen global-context feature extraction, a C3k2_EMA module to enhance localization of small and occluded targets, and Poly-Scale Convolution (PSConv) to preserve structural details while reducing computational redundancy. In addition, a lightweight attention mechanism is introduced to improve feature discrimination in cluttered backgrounds. Experimental results on both self-constructed and generalization datasets show that the proposed method achieves a precision of 92.6%, a recall of 86.8%, and an mAP50 of 92.1% with only 7.26 M parameters. The framework also demonstrates strong robustness and generalization across different chili varieties. When deployed on an edge computing platform (NVIDIA Jetson AGX Orin), the model achieves real-time inference at 39.5 FPS. Furthermore, validation experiments under controlled indoor conditions show that the proposed framework can effectively support simulated pollination tasks, achieving a success rate of 90.0% for upwardfacing flowers. These results indicate that CFPR-YOLO provides an effective visual perception solution for agricultural engineering informatics-oriented pollination systems and offers practical potential for precision pollination and intelligent fruit-set management in horticultural production.
Why it matches plant phenotyping methodsチリ花の検出・姿勢推定という植物器官の画像計測手法を開発し、データセット、汎化性能、エッジ実装、実環境に近い条件での検証まで行っており、単なる受粉実験の補助計測ではない。
abstractthis study proposes a lightweight and robust edge vision framework, termed CFPR-YOLO, for chili flower detection and pose-aware perception under complex agricultural conditions.
GrapevineNeRF / 3D Gaussian SplattingLiDAR / point cloudRGB / grayscaleFruitStem / branchPose / keypoint estimation2D/3D reconstructionSegmentation
• End-to-end pipeline from neural reconstruction to physical grape berry manipulation. • Efficient point clouds generation using NeRF and the metric scale derived directly from robot kinematics. • RANSAC sphere fitting achieves 92.1% berry detection precision without annotated training data. • Stem-aligned 6-DoF pose optimization improves end-to-end grip success by 17.2%. Table grape thinning requires selective removal of 20–40% of berries from dense clusters. In practice, workers decide which berries to remove by considering both the approximate berry count and local 3D spatial characteristics such as crowding and relative positioning. Automating this task is challenging because conventional 2D image-based approaches suffer from occlusion-related counting errors and lack explicit 3D spatial information necessary for reliable manipulation. We propose a robot-integrated vision pipeline that reconstructs grape bunch structure from posed multi-view RGB images. Neural Radiance Fields (NeRF) is used to learn a volumetric scene representation, from which a dense, low-noise point cloud is extracted via depth back-projection, and RANSAC-based geometric fitting models individual berries and stems, enabling berry-level segmentation and orientation estimation for manipulation planning. The perception pipeline uses an eye-in-hand RealSense D405 camera mounted on a Fanuc CRX-5iA collaborative robot. Camera poses are derived from the robot kinematic chain, allowing the reconstructed point cloud and detected berry centers to be expressed directly in the metric robot base frame without external scale recovery. In robot-mounted RealSense D405 experiments on 10 grape bunches, RANSAC sphere fitting achieved a counting MAE of 0.50 berries, RMSE of 0.71 berries, and mean center localization error of 2.71 mm. On a 52-bunch benchmark, RANSAC sphere fitting outperforms the learning-based SoftGroup++ method for 3D berry instance segmentation (92.1% vs 82.8% average precision) without requiring annotated training data. In 35 manipulation trials, the system achieved an 85.7% pre-grasp reachability rate and an 83.3% conditional target success rate demonstrating an end-to-end pipeline from neural reconstruction to manipulation-ready berry poses.
Why it matches plant phenotyping methodsブドウ房の3D再構成、ベリー分割・計数・位置推定を中核とするロボット統合型フェノタイピング手法であり、技術性能も定量評価している。
abstractWe propose a robot-integrated vision pipeline that reconstructs grape bunch structure from posed multi-view RGB images.
To address the low efficiency of manual inspection for corn sowing quality, which is labor-intensive and time-consuming, this study proposes an automatic seedling-stage plant-spacing measurement method based on three-dimensional machine vision. A high-clearance mobile platform equipped with a ZED 2i stereo camera and an industrial computer was developed to acquire RGB images and depth information of corn seedlings in the field in real time. Using the YOLOv11-Pose model, plant keypoints were detected and localized; combined with camera calibration and 3D reconstruction techniques, inter-plant distances were computed automatically. A sowing-quality evaluation framework was then established to enable automated analysis of the Quality of Feed Index (QFI), Multiple Index (MUL), Miss Index (MI), and coefficient of variation. Experimental results indicate that the system operates effectively under three preset plant spacings (15 cm, 20 cm, and 25 cm), achieving a keypoint-detection mAP@0.5 of 0.990 and an mAP@0.5:0.95 of 0.989. In sowing-quality evaluation, the qualified indices produced by the system were 77.83%, 80.36%, and 82.46%, respectively, and the Quality of Feed Index (QFI), Multiple Index (MUL), Miss Index (MI), and coefficient of variation showed trends consistent with manual measurements. The proposed method enables efficient, nondestructive detection of seedling-stage plant spacing and sowing quality, providing reliable technical support for precision sowing and field management.
Why it matches plant phenotyping methods3Dマシンビジョン、姿勢検出、校正、3D再構成を組み合わせ、トウモロコシ個体間距離を自動推定・検証する手法が研究の中心である。
abstractthis study proposes an automatic seedling-stage plant-spacing measurement method based on three-dimensional machine vision.
Plant phenotyping relevance match · UnverifiedOpenAlex · Europe PMC · Crossref · checked 5 Sept 2026
WheatNeRF / 3D Gaussian SplattingLiDAR / point cloudWhole plant / canopy / plot / fieldPose / keypoint estimation2D/3D reconstruction
Precise reconstruction of plant phenotypes is crucial for smart agriculture. Conventional methods struggle with low efficiency and strong dependency on high-quality data, especially for low-texture and structurally complex crops like wheat. We propose a novel 3D reconstruction framework—Plant3R—that fuses deep feature learning with 3D Gaussian Splatting (3DGS). It innovatively uses the Matching and Stereo 3D Reconstruction (MASt3R) model for sparse point cloud reconstruction and camera pose estimation via its 3D feature matching capabilities, which substantially improve image matching rates and the quality of sparse point clouds. Subsequently, 3DGS is employed for rendering and optimization, enabling end-to-end, high-fidelity, and high-robust 3D reconstruction of wheat plants. Validated on potted wheat at multiple growth stages using handheld images, our experimental results demonstrate that Plant3R performs well in feature extraction and matching, and the reconstructed point cloud provides a good geometric prior for the subsequent rendering stage. In most scenes, its key rendering metrics—Peak Signal-to-Noise Ratio (PSNR) > 34, Structural Similarity Index Measure (SSIM) of 0.94, and Learned Perceptual Image Patch Similarity (LPIPS) 0.94), confirming its utility for accurate and quantitative phenotype analysis. Overall, Plant3R not only improves the rendering quality and geometric precision of 3D modeling, but also provides a reliable tool for accurate phenotypic parameter extraction and high-throughput crop phenotyping in precision agriculture.
Why it matches plant phenotyping methods小麦植物の3D再構成と表現型パラメータ抽出を目的とする画像解析手法を開発・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。
abstractWe propose a novel 3D reconstruction framework—Plant3R—that fuses deep feature learning with 3D Gaussian Splatting (3DGS).
MilletPeaField / plotGreenhouseNeRF / 3D Gaussian SplattingWhole plant / canopy / plot / fieldPose / keypoint estimation2D/3D reconstructionVisualization / data managementGrowth / development / phenology
Plant phenotyping in precision agriculture increasingly requires high-fidelity three-dimensional reconstruction and accessible visualization methods. This study presents an integrated pipeline combining Neural Radiance Fields (NeRF), 3D Gaussian Splatting (G-Splat), and Virtual Reality (VR) visualization for comprehensive plant analysis across developmental stages. We collected multi-view imagery of finger millet, proso millet, mungbean, and field pea under controlled greenhouse conditions, aligning data acquisition with standardized BBCH phenological scales. Camera pose estimation was performed using GLOMAP, followed by reconstruction via both Nerfacto and G-Splat implementations. Quantitative evaluation using PSNR, SSIM, and LPIPS metrics revealed complementary strengths of the two approaches: G-Splat achieved superior structural fidelity, while NeRF provided enhanced perceptual realism. Both reconstruction methods were successfully integrated into an immersive VR greenhouse environment deployed on Meta Quest headsets, maintaining consistently high framerates. This framework establishes a practical foundation for incorporating neural reconstruction and immersive technologies into agricultural phenotyping workflows, supporting both research applications and educational engagement.
Why it matches plant phenotyping methods植物の多視点画像からNeRFと3D Gaussian Splattingで3D形状を再構成し、画質指標で比較評価する統合フェノタイピング基盤の開発・検証が中心である。
abstractThis study presents an integrated pipeline combining Neural Radiance Fields (NeRF), 3D Gaussian Splatting (G-Splat), and Virtual Reality (VR) visualization for comprehensive plant analysis across developmental stages.
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.
Efficient large-scale 3D reconstruction of orchard environments is essential for robotic inspection and precision agriculture, yet existing methods struggle with unstructured scenes, variable illumination, and computational bottlenecks. We propose InspectGaussian, a coarse-to-fine Gaussian reconstruction framework tailored for orchard inspection robots. The pipeline integrates an RGB-D-based data acquisition strategy using ORB-SLAM3, which is enhanced by a dense mapping module for robust large-scale pose estimation and point cloud generation. A divide-and-conquer strategy is then employed: individual plant views are extracted via a YOLO-World-based detection and 3D matching algorithm, followed by plant-specific reconstruction using an improved 3D Gaussian Splatting (3DGS) method incorporating depth regularization and region-aware refinement. Experimental results in citrus orchards demonstrate that InspectGaussian achieves 96% average precision and 93% recall in plant view extraction, while surpassing state-of-the-art methods in reconstruction fidelity (31.226 PSNR, 0.915 SSIM, 0.067 LPIPS) and point cloud accuracy (7 mm error). These results confirm its effectiveness in capturing fine structural and textural details while maintaining scalability and efficiency. This framework provides a practical solution for high-throughput, in-field plant phenotyping and lays the foundation for intelligent orchard monitoring and management.
Why it matches plant phenotyping methods植物個体の3D再構成とRGB-D・検出・Gaussian Splattingを統合した手法開発であり、植物の構造的形質取得を目的とするため、フェノタイピング手法が中心である。
abstractWe propose InspectGaussian, a coarse-to-fine Gaussian reconstruction framework tailored for orchard inspection robots.
Reproduction assets foundThe paper's authors explicitly state their analysis code is publicly available on GitHub. Phenotype datasets (RGB-D orchard image sequences, LiDAR point clouds, manual trait measurements) are only available upon request, so they do not qualify as public assets.Code · publicOur code are available at https://github.com/zlhzau/InspectGaussian.git .Open asset ↗zlhzau/InspectGaussianlines:489-515Plant phenotyping relevance match · UnverifiedEurope PMC · checked 5 Sept 2026
Intelligent recognition and rapid grading of tea buds are crucial for advancing tea-picking machinery; however, complex plantation backgrounds and inconsistent bud growth have limited traditional algorithms to merely identifying picking points, neglecting bud pose and grade, which restricts harvesting efficiency. To address these challenges, we propose YOLO-PC, a deep neural network designed for simultaneous tea bud pose estimation and classification, which incorporates a dynamic snake convolution (DSConv) module for enhanced shape feature extraction, an ELASPP-CSPC attention mechanism for improved spatial pooling, and EIoU loss to accelerate regression and boost localization accuracy. Experimental results demonstrate that the model achieves detection accuracies of 91.5% for one-bud-one-leaf and 93.2% for one-bud-two-leaf scenarios, with an average keypoint detection accuracy (Pose_mAP) of 89.7% and a Normalized Mean Error (NME) of 0.047; furthermore, compared to YOLOv7-pose, it increases mean average precision by 7.26% and pose accuracy by 9.65% while reducing parameters by 14.99 M. Ablation studies confirm the superior performance of the proposed model in tea bud detection, indicating its potential to provide robust practical support for adaptive and intelligent tea harvesting systems.
Why it matches plant phenotyping methods茶芽の姿勢・等級という植物器官の状態を画像から推定する深層学習手法を開発し、精度比較・アブレーション評価まで行っており、単なる収穫対象の位置検出を超えたフェノタイピング手法が中心である。
abstractwe propose YOLO-PC, a deep neural network designed for simultaneous tea bud pose estimation and classification
Robotic pollination represents a pivotal component of smart agriculture, with foundational architectures for target recognition, path planning, and motion control having been progressively established. However, developing an efficient and robust pollination system that integrates perception, decision-making, and execution within real-world scenarios remains confronted with complex challenges. This study systematically reviews recent advancements in the field and distills the core technical issues of greenhouse robotic pollination into three primary domains: target detection and pose estimation, end-effector design, and pollination strategies combined with motion control. Focusing on the visual perception of flowers, actuator architecture, and operational tactics, this review synthesizes existing academic findings to evaluate the state-of-the-art in flower detection and pose estimation, characterize diverse end-effector designs, and analyze the evolutionary trajectory of motion control techniques. Specifically, the analysis encompasses the impact of detection algorithms on recognition accuracy and robustness, the structural classification and performance attributes of pollination mechanisms, and the optimization of control strategies. Furthermore, the study categorizes global research backgrounds, technical methodologies, and paradigmatic system cases, offering a critical evaluation of experiences in constructing automated pollination systems. Despite these advances, current robotic pollination technologies for peppers (chili) face significant bottlenecks characterized by immature methods for precise flower detection and pose estimation, the need for optimized specialized end-effector designs, and insufficient robustness in decision-making systems under dynamic environmental conditions. To address these issues, future development should prioritize constructing diverse, large-scale flower image and pose datasets while developing detection algorithms adaptable to complex environments to achieve high-precision identification. Additionally, implementing this system requires a hierarchical architecture where perception drives adaptive actuation. Deep learning models must localize flower targets and assess maturity in real-time, feeding coordinates to path planners that generate collision-free trajectories through foliage. These trajectories are executed via multimodal motion control, synchronizing the rigid manipulator with soft end-effectors. By embedding tactile feedback into the machine learning loop, the system creates a unified sensorimotor framework. This enables dynamic force modulation based on physical resistance, ensuring precise, non-destructive pollination tailored to chili plants.
Why it matches plant phenotyping methods温室コショウの花の検出・姿勢推定など、植物器官の観測・形質抽出を中核とするロボット受粉技術のレビューであり、方法論的貢献が中心。
abstractThis study systematically reviews recent advancements in the field and distills the core technical issues of greenhouse robotic pollination into three primary domains: target detection and pose estimation, end-effector design, and pollination strategies combined with motion control.
• 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.
Accurate localization of plant growth points is essential for precision agriculture applications, including electro-weeding and laser weeding. While crop and weed detection has been extensively studied, existing methods focus primarily on object-level recognition and often neglect fine-grained growth point localization. To address this limitation, we propose a novel training strategy, epoch-based prior annealing (EPA), which incorporates the excess green minus excess red (ExG-ExR) index as prior knowledge and introduces schedule factor and gain factor to effectively steer keypoint regression. The experimental results show that incorporating EPA improves keypoint localization performance, with mAP50 increasing by 0.024 and mAP50:95 by 0.011, while maintaining bounding box detection performance. The parameter sensitivity experiments confirmed that both excessively strong and weak guidance can hinder training. Furthermore, analysis of parameters and computational cost shows that the additional overhead introduced by the EPA strategy accounts for less than 0.5% of the total, and be considered negligible. In summary, the proposed EPA strategy significantly improves the accuracy, robustness, and generalizability of plant and growth point detection models, offering a practical and scalable solution for precision agricultural applications.
Why it matches plant phenotyping methods植物の生長点を画像から局在化する学習戦略を開発し、その性能を実験的に検証しており、植物器官の表現型取得が中心です。
abstractwe propose a novel training strategy, epoch-based prior annealing (EPA), which incorporates the excess green minus excess red (ExG-ExR) index as prior knowledge and introduces schedule factor and gain factor to effectively steer keypoint regression.
Reproduction assets found保存済みの本文根拠を更新済みルールで再検証し、公開資産1件を確認しました。Dataset · publicYou can access the dataset used in this study at the following links: https://github.com/cropandweed/cropandweed-dataset.Open asset ↗cropandweed/cropandweed-datasethtml-lines:497-528Plant phenotyping relevance match · UnverifiedCrossref · OpenAlex · checked 15 Sept 2026
Published1 Jan 2026ISPRS Journal of Photogrammetry and Remote SensingCited by 6 · OpenAlex ↗
Three-dimensional (3D) plant modeling is fundamental for precise phenotyping analysis. In this study, a high-throughput, multi-stage 3D reconstruction pipeline is proposed to efficiently generate point clouds with real-world physical scales. The pipeline comprises five core components: data acquisition, semantic segmentation, sparse reconstruction, dense reconstruction, and phenotypic trait extraction. To enhance the accuracy of plant structure identification, the SegFormer semantic segmentation model is employed for pixel-level segmentation, thereby guiding the subsequent reconstruction stages to focus specifically on plant regions. In feature-based sparse reconstruction, the scarcity of texture information often results in an insufficient number of matching point pairs, leading to failures in camera parameter estimation. Furthermore, the reconstructed point clouds frequently lack consistency with real-world scale. To address these challenges, a calibration-constrained sparse reconstruction method, Sparse Reconstruction from Calibrated Images (SRCI) was proposed. By integrating precise calibration results computed in a custom world coordinate system, SRCI circumvents the limitations of traditional feature matching in scenarios with scarce features, thereby resolving camera pose estimation failures caused by insufficient matching pairs and generating sparse point clouds with true physical scale. Subsequently, CL-MVSNet was employed to generate dense point clouds. The validation experiments are conducted from three perspectives: visual comparison, phenotypic accuracy assessment, and reconstruction accuracy evaluation. First, five groups of rapeseed plants are selected to perform visual comparisons between the proposed reconstruction pipeline and other advanced reconstruction software. The results demonstrate that the proposed reconstruction pipeline achieves superior performance in terms of visual quality. Additionally, three phenotypic parameters of rapeseed plants—plant height, leaf width, and chord length are manually measured and compared with the corresponding phenotypic parameters extracted from the reconstructed point clouds. The analysis revealed mean absolute errors of 4.93 mm, 3.16 mm, and 6.02 mm; root mean square errors of 6.38 mm, 4.56 mm, and 8.35 mm; and coefficients of determination of 0.98, 0.94, and 0.93, respectively. To further validate the generalization performance and accuracy of the proposed method, four additional plant categories with progressively increasing complexity were selected for accuracy evaluation. For the first three plant categories, the Chamfer distances between the reconstructed point clouds and ground truth point clouds were all within 2.4 mm. In the most complex rice reconstruction experiments, the Chamfer distances between the reconstructed point clouds and ground truth point clouds were all within 9 mm, while other methods failed to achieve effective reconstruction. The proposed high-throughput, automated, and high-quality 3D reconstruction framework provides reliable technical support and data resources for genetic research applications, including gene localization, quantitative trait locus analysis, and genome-wide association studies. • We propose a Plant-to-Camera system for high-quality 3D plant reconstruction within 6 minutes, showing strong generalizability. • Our Sparse Reconstruction from Calibrated Images (SRCI) method prevents failures in feature-scarce scenes. • We develop SFNet, a feature descriptor module that fuses multi-frequency to enhance plant feature representation.
Why it matches plant phenotyping methods植物の3D再構成と形質抽出パイプラインを開発し、植物形質および再構成精度を検証しており、フェノタイピング手法が中心である。
abstracta high-throughput, multi-stage 3D reconstruction pipeline is proposed to efficiently generate point clouds with real-world physical scales
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.
Addressing the challenges of variable target morphology, small critical regions, and complex background interference in eggplant picking point detection within complex agricultural scenarios, this study proposes MDAD-YOLO (Multi-dimensional Attention and DySample YOLO), a detection model improved based on the YOLOv10n-pose framework. First, the model’s cross-dimensional perception ability for fruits and picking points is enhanced by integrating the collaborative mechanism of regional receptive field attention with channel-space joint attention. Next, within the Neck structure, coordinate attention is incorporated to optimize the spatial localization accuracy of fine-grained features, enhancing sensitivity to minute regions such as the fruit stem apex. Additionally, dynamic pixel reorganization is applied to enhance feature map reconstruction details, addressing the detail loss caused by traditional interpolation methods. Finally, cascading adaptive fine-grained channel attention with position-sensitive attention enables multi-level modeling of channel dependencies and collaborative spatial context enhancement. Through a seven-tier validation framework, the model’s effectiveness, robustness, and generalizability have been comprehensively demonstrated. Experimental results show that the model achieves 93.6% mAP@50 for object detection, 94.7% mAP@50 and 92.1% mAP for keypoints detection, and an average pixel Euclidean distance error of 19.41 on the self-built eggplant dataset, outperforming YOLOv12 and other high-performance models. Additionally, cross-crop experiments on the pepper dataset showed a 2.1% and 2.7% improvement in mAP for object and picking point detection, respectively, compared to the baseline model, confirming its cross-crop robustness. This study reveals the synergistic enhancement of dynamic upsampling and attention mechanisms in agricultural object detection, providing new insights for lightweight model design in complex scenarios.
Why it matches plant phenotyping methods果実と収穫点の画像ベース検出・キーポイント推定モデルを開発し、複数データセットで性能と頑健性を検証しているため、植物形質取得手法が中心である。
abstractthis study proposes MDAD-YOLO (Multi-dimensional Attention and DySample YOLO), a detection model improved based on the YOLOv10n-pose framework.
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 $\leq95\%$ 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 methodsリンゴの姿勢推定アノテーションを大幅に効率化する3D再構成・自動ラベル投影パイプラインを開発し、姿勢推定性能も評価しているため、植物フェノタイピング手法が中心である。
abstractWe 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.
Reproduction assets foundThe paper explicitly provides two paper-specific public assets: the authors' phenotyping/pose-estimation pipeline code on GitHub and the collected apple orchard image dataset on a 4TU DOI. Both are directly used for the paper's measurements and analysis.Dataset · publicIn total, 367 images were collected. The dataset is available at https://doi.org/10.4121/976c94f2-028f-4291-adfd-20eb82b0f647Open asset ↗10.4121/976c94f2-028f-4291-adfd-20eb82b0f647lines:92-108Plant phenotyping relevance match · UnverifiedCrossref · checked 15 Sept 2026
The synergistic development of Internet of Things(IoT), robotics, and Artificial Intelligence (AI) is reshaping the technological paradigms of interdisciplinary laboratories and industrial ecosystems. IoT-enabled vertical farming systems demonstrate significant advantages, achieving yield enhancement while reducing carbon emissions compared to traditional agriculture, thereby providing innovative solutions for sustainable food production. The advancement of robotic technologies further expands the application dimensions of mobile intelligent sensors in vertical farm IoT networks. Based on an autonomous farming system that integrates Unmanned Aerial Vehicle (UAV), sensors, and modular vertical farming units, this study proposes a three dimensional Scene Graph (3DSG)-based hierarchical mapping method for the dynamic monitoring of plant and fruit growth. Through feedback mechanisms, the system optimizes growth conditions by adjusting lighting and nutrient delivery, while the hierarchical mapping architecture reduces detection errors and enables comprehensive 3D visualization. The main contributions of this research include: 1) Pioneering application of 3DSG technology to establish a multi-dimensional spatiotemporal representation model for plant growth processes, supporting interpretable analysis and traceable monitoring; 2) Establishing an uncertainty model through error propagation by systematically analyzing sensor models (covering various common sensor combinations) and integrating these models into 3D object pose estimation algorithms. This highlights the necessity of hierarchical abstraction levels. The system is validated through simulations and real-world experiments, providing a quantitative evaluation of object pose estimation; and 3) An IoT-driven intelligent vertical farming architecture that integrating mobile robotic perception networks and environmental regulation devices, enabling dynamic acquisition and closed-loop control of plant growth parameters. Open-source code is available at https://github.com/allenthreee/scene_graph, video link: https://youtu.be/dhc8RLmX7hc.
Why it matches plant phenotyping methods植物・果実の成長監視と計数を目的に、3Dシーングラフ、SLAM、センサー融合、誤差伝播モデルを開発・検証しており、表現型取得手法が研究の中心である。
titleHierarchical 3-D Scene-Graph-Based Semantic-Metric SLAM for Plant Inspection and Fruit Counting in Intelligent Hydroponics System
CherryTomatoGreenhouseNeRF / 3D Gaussian SplattingPhotogrammetry / SfM / MVSLiDAR / point cloudFruitPose / keypoint estimation2D/3D reconstructionSegmentation
Accurate and nondestructive monitoring of tomato growth is essential for large-scale greenhouse production; however, it remains challenging for small-fruited cultivars such as cherry tomatoes. Traditional 2D image analysis often fails to capture precise morphological traits, limiting its usefulness in growth modeling and yield estimation. This study proposes an automated phenotyping framework that integrates deep learning-based instance segmentation with high-resolution 3D point cloud reconstruction and ellipsoid fitting to estimate fruit size and ripeness from daily video recordings. These techniques enable accurate camera pose estimation and dense geometric reconstruction (via SfM and MVS), while Nerfacto enhances surface continuity and photorealistic fidelity, resulting in highly precise and visually consistent 3D representations. The reconstructed models are followed by CIELAB color analysis and logistic curve fitting to characterize the growth dynamics. When applied to real greenhouse conditions, the method achieved an average size estimation error of 8.01% compared to manual caliper measurements. During summer, the maximum growth rate (gmax) of size and ripeness were 24.14%, and 95.24% higher than in winter, respectively. Seasonal analysis revealed that winter-grown tomatoes matured approximately 10 days later than summer-grown fruits, highlighting environmental influences on phenological development. By enabling precise, noninvasive tracking of size and ripeness progression, this approach is a novel tool for smart and sustainable agriculture.
Why it matches plant phenotyping methods深層学習によるインスタンスセグメンテーション、3D再構成、色解析を統合し、トマト果実のサイズと成熟度を推定するフェノタイピング手法の開発・評価が中心である。
abstractThis study proposes an automated phenotyping framework that integrates deep learning-based instance segmentation with high-resolution 3D point cloud reconstruction and ellipsoid fitting to estimate fruit size and ripeness from daily video recordings.
This study introduces a versatile seed 3D reconstruction method that is applicable to multiple crops—including maize, wheat, and rice—and designed to overcome the inefficiency and subjectivity of manual measurements and the high costs of laser-based phenotyping. A panoramic video of the seed is captured and processed through frame sampling to extract multi-view images. Structure-from-Motion (SFM) is employed for sparse reconstruction and camera pose estimation, while 3D Gaussian Splatting (3DGS) is utilized for high-fidelity dense reconstruction, generating detailed point cloud models. The subsequent point cloud preprocessing, filtering, and segmentation enable the extraction of key phenotypic parameters, including length, width, height, surface area, and volume. The experimental evaluations demonstrated a high measurement accuracy, with coefficients of determination (R2) for length, width, and height reaching 0.9361, 0.8889, and 0.946, respectively. Moreover, the reconstructed models exhibit superior image quality, with peak signal-to-noise ratio (PSNR) values consistently ranging from 35 to 37 dB, underscoring the robustness of 3DGS in preserving fine structural details. Compared to conventional multi-view stereo (MVS) techniques, the proposed method can achieve significantly improved reconstruction accuracy and visual fidelity. The key outcomes of this study confirm that the 3DGS-based pipeline provides a highly accurate, efficient, and scalable solution for digital phenotyping, establishing a robust foundation for its application across diverse crop species.
Why it matches plant phenotyping methods3DGSを用いた種子の3D再構成・点群処理・形質抽出パイプラインを開発し、精度を評価しており、植物表現型取得手法が研究の中心である。
abstractThis study introduces a versatile seed 3D reconstruction method that is applicable to multiple crops
StrawberryField / plotLiDAR / point cloudFlowerObject detectionPose / keypoint estimation
The small scale of urban farms and the commercial availability of low-cost robots (such as the FarmBot) that automate simple tending tasks enable an accessible platform for plant phenotyping. We have used a FarmBot with a custom camera end-effector to estimate strawberry plant flower pose (for robotic pollination) from acquired 3D point cloud models. We describe a novel algorithm that translates individual occupancy grids along orthogonal axes of a point cloud to obtain 2D images corresponding to the six viewpoints. For each image, 2D object detection models for flowers are used to identify 2D bounding boxes which can be converted into the 3D space to extract flower point clouds. Pose estimation is performed by fitting three shapes (superellipsoids, paraboloids and planes) to the flower point clouds and compared with manually labeled ground truth. Our method successfully finds approximately 80% of flowers scanned using our customized FarmBot platform and has a mean flower pose error of 7.7 degrees, which is sufficient for robotic pollination and rivals previous results. All code will be made available at https://github.com/harshmuriki/flowerPose.git.
Why it matches plant phenotyping methodsカスタムカメラ付きロボットによる3D花姿勢推定アルゴリズムとプラットフォームを開発・評価しており、花の姿勢という植物形質の取得が中心である。
abstractenable an accessible platform for plant phenotyping
Reproduction assets foundThe paper's flower pose estimation pipeline (translating occupancy grid, 2D/3D conversion, shape fitting) has an explicit authors' code deposit statement with a public GitHub URL, phrased as future availability ('will be made available'), so actionability is likely but not fully confirmed. No public dataset of the FarmCode · publiclower point clouds and compared with manually labeled ground truth. Our method successfully finds approximately 80% of flowers scanned using our customized FarmBot platform and has a mean flower pose error of 7.7 degrees, which is sufficient for robotic pollination and rivals previous results. All code will be made available at https://github.com/harshmuriki/flowerPose.git .
I Introduction
Urban farms [ 1 ] provide healthy food to local communities and can serve as platforms for education and sustainability. Unlike their rural counterparts, urban farms are usually small in scale and commercially available robotic systems such as the FarmBot [ 2 ] have been developed to help automate basic cuOpen asset ↗harshmuriki/flowerPoselines:1-53Plant phenotyping relevance match · UnverifiedEurope PMC · checked 6 Sept 2026
Background Accurate sorghum spike detection is critical for monitoring growth conditions, accurately predicting yield, and ensuring food security. Deep learning models have improved the accuracy of spike detection thanks to advances in artificial intelligence. However, the dense distribution of sorghum spikes, variable sizes and complex background information in UAV images make detection and counting difficult. Methods We propose a multiscale and oriented sorghum spike detection and counting model in UAV images (MOSSNet). The model creates a Deformable Convolution Spatial Attention (DCSA) module to improve the network's ability to capture small sorghum spike features. It also integrated Circular Smooth Labels (CSL) to effectively represent morphological features. The model also employs a Wise IoU-based localization loss function to improve network loss. Results Results show that MOSSNet accurately counts sorghum spike under field conditions, achieving mAP of 90.3%. MOSSNet shows excellent performance in predicting spike orientation, with RMSEa and MAEa of 14.6 and 12.5 respectively, outperforming other directional detection algorithms. Compared to general object detection algorithms which output horizonal detection boxes, MOSSNet also demonstrates high efficiency in counting sorghum spikes, with RMSE and MAE values of 9.3 and 8.1, respectively. Discussion Sorghum spikes have a slender morphology and their orientation angles tend to be highly variable in natural environments. MOSSNet 's ability has been proved to handle complex scenes with dense distribution, strong occlusion, and complicated background information. This highlights its robustness and generalizability, making it an effective tool for sorghum spike detection and counting. In the future, we plan to further explore the detection capabilities of MOSSNet at different stages of sorghum growth. This will involve implementing object model improvements tailored to each stage and developing a real-time workflow for accurate sorghum spike detection and counting.
Why it matches plant phenotyping methodsUAV画像からソルガム穂の検出・計数・向き推定を行う深層学習手法を開発し、精度評価も実施しており、植物形態形質の取得手法が中心である。
abstractWe propose a multiscale and oriented sorghum spike detection and counting model in UAV images (MOSSNet).
For fruit harvesting robots, rapid and accurate detection of fruits and picking points is one of the main challenges for their practical deployment. Several fruits typically grow in clusters or bunches, such as grapes, cherry tomatoes, and blueberries. For such clustered fruits, it is desired for them to be picked by bunches instead of individually. This study proposes utilizing a low-cost off-the-shelf RGB-D sensor mounted on the end effector and a lightweight improved YOLOv8-Pose neural network to detect cherry tomato bunches and picking points for robotic harvesting. The problem of occlusion and overlap is alleviated by merging RGB and depth images from the RGB-D sensor. To enhance detection robustness in complex backgrounds and reduce the complexity of the model, the Starblock module from StarNet and the coordinate attention mechanism are incorporated into the YOLOv8-Pose network, termed StarBL-YOLO, to improve the efficiency of feature extraction and reinforce spatial information. Additionally, we replaced the original OKS loss function with the L1 loss function for keypoint loss calculation, which improves the accuracy in picking points localization. The proposed method has been evaluated on a dataset with 843 cherry tomato RGB-D image pairs acquired by a harvesting robot at a commercial greenhouse farm. Experimental results demonstrate that the proposed StarBL-YOLO model achieves a 12% reduction in model parameters compared to the original YOLOv8-Pose while improving detection accuracy for cherry tomato bunches and picking points. Specifically, the model shows significant improvements across all metrics: for computational efficiency, model size (−11.60%) and GFLOPs (−7.23%); for pickable bunch detection, mAP50 (+4.4%) and mAP50-95 (+4.7%); for non-pickable bunch detection, mAP50 (+8.0%) and mAP50-95 (+6.2%); and for picking point detection, mAP50 (+4.3%), mAP50-95 (+4.6%), and RMSE (−23.98%). These results validate that StarBL-YOLO substantially enhances detection accuracy for cherry tomato bunches and picking points while improving computational efficiency, which is valuable for resource-constrained edge-computing deployment for harvesting robots.
Why it matches plant phenotyping methodsRGB-D画像と改良YOLOv8-Poseを用いて、収穫対象となるトマト房と摘採点を検出・位置推定する手法を開発し、実データセットで性能検証している。単なる収穫対象の位置検出に見えるが、房の可収穫性と摘採点という植物器官の状態・位置を抽出する技術的貢献が中心である。
abstractThis study proposes utilizing a low-cost off-the-shelf RGB-D sensor mounted on the end effector and a lightweight improved YOLOv8-Pose neural network to detect cherry tomato bunches and picking points for robotic harvesting.
Flag leaf angle (FLANG) is one of the key traits in wheat breeding due to its impact on plant architecture, light interception, and yield potential. An image-based method of measuring FLANG in wheat would reduce the labor and error of manual measurement of this trait. We describe a method for acquiring in-field FLANG images and a lightweight deep learning model named LeafPoseNet that incorporates a spatial attention mechanism for FLANG estimation. In a test dataset with wheat varieties exhibiting diverse FLANG, LeafPoseNet achieved high accuracy in predicting the FLANG, with a mean absolute error (MAE) of 1.75°, a root mean square error (RMSE) of 2.17°, and a coefficient of determination ( R 2 ) of 0.998, significantly outperforming established models such as YOLO12x-pose, YOLO11x-pose, HigherHRNet, Lightweight-OpenPose, and LitePose. We performed phenotyping and genome-wide association study to identify the genomic regions associated with FLANG in a panel of 221 diverse bread wheat genotypes, and identified 10 quantitative trait loci. Among them, qFLANG2B.2 was found to harbor a potential causal gene, TraesCS2B01G313700 , which may regulate FLANG formation by modulating brassinosteroid levels. This method provides a low-cost, high-accuracy solution for in-field phenotyping of wheat FLANG, facilitating both wheat FLANG genetic studies and ideal plant type breeding.
Why it matches plant phenotyping methods小麦の旗葉角度を画像から推定する手法と深層学習モデルを開発し、既存モデルとの精度比較・検証および圃場フェノタイピングに適用しており、表現型取得法が研究の中心である。
abstractAn image-based method of measuring FLANG in wheat would reduce the labor and error of manual measurement of this trait.
To address the issues of low efficiency and large errors in traditional manual cucumber fruit phenotyping methods, this paper proposes the application of keypoint detection technology for cucumber phenotyping and designs an improved lightweight model called YOLO11m-SCFPose. Based on YOLO11m-pose, the original backbone network is replaced with the lightweight StarNet-S1 backbone, reducing model complexity. Additionally, an improved C3K2_PartialConv neck module is used to enhance information interaction and fusion among multi-scale features while maintaining computational efficiency. The Focaler-IoU loss function is employed to improve keypoint localization accuracy. Results show that the improved model achieves an mAP50-95 of 0.924, with a floating-point operation count (GFLOPs) of 32.1, and reduces the model size to 1.229 × 107 parameters. This model demonstrates better computational efficiency and lower resource consumption, providing an effective lightweight solution for crop phenotypic analysis.
Why it matches plant phenotyping methodsキュウリ果実の表現型取得を目的に、キーポイント検出モデルを設計・改良し、精度と計算効率を評価しているため、植物フェノタイピング手法が中心である。
abstractthis paper proposes the application of keypoint detection technology for cucumber phenotyping and designs an improved lightweight model called YOLO11m-SCFPose.
Leaf shape is of great significance in plant phenotype research. Landmarks method is a widely used morphometric approach, which can comprehensively describe the morphological differences among leaves. However, the selection of landmarks is time-consuming and laborious. An automatic landmarking algorithm is proposed here. Based on conformal mapping, the leaf outline can be transformed into a monotonically increasing function curve, referred to as the ’fingerprint function’. The Dynamic Time Warping (DTW) algorithm was introduced to match landmarks between different leaves. Two leaf datasets were used to validate the algorithm separately in different species and developmental stages. Dataset1 is a public dataset which covers 26 different types of leaves. The average positional difference between automatic and manual landmarks for dataset1 was only 2.95%. Dataset2 consists of cotton leaves collected in the field at various growth stages, and the positional difference for this dataset was all below 5%. These results validate that our algorithm is applicable to a wide range of leaf types and capable of identifying and locating novel features that emerge during leaf growth. The automatic landmarking algorithm can simulate manual landmarking to a great extent. It provides a new approach for automated acquisition of plant leaf shape homology tailored to the research needs of botanists.
Why it matches plant phenotyping methods葉形態のランドマークを自動抽出する手法を開発し、複数の葉データセットで精度検証しているため、植物表現型取得法が研究の中心です。
abstractAn automatic landmarking algorithm is proposed here.
Plant phenotyping relevance match · UnverifiedOpenAlex · Europe PMC · checked 6 Sept 2026
BACKGROUND: Phenotypic characterization of mature soybean pods is a crucial aspect of breeding programs, yet efficiently obtaining accurate pod phenotypic parameters remains a major challenge. Recent advances in deep learning, particularly in keypoint detection models, have introduced innovative methods for pod phenotype extraction. However, precise identification and analysis of fine-scale phenotypic traits in soybean pods remain challenging in current research. RESULTS: We propose Pod-pose, an innovative top-down keypoint detection model for precise soybean pod phenotyping that adapts human pose estimation techniques to plant phenotyping. Specifically, Pod-pose integrates the architectural strengths of various advanced YOLO (You Only Look Once) models through bottleneck structure optimization and positional feature enhancement to achieve superior detection accuracy. Furthermore, we implemented a two-stage detection method augmented with transfer learning, which not only reduces training complexity but also significantly enhances the model's performance. Extensive evaluation of our custom-built dataset demonstrated Pod-Pose's superior performance, with the X variant achieving an Average Precision of 0.912 at an IoU threshold of 0.5 (AP@IoU = 0.5). Notably, four critical pod-related phenotypic traits were successfully quantified: pod length, bending length, curvature, and inflection point width. CONCLUSIONS: This study establishes Pod-Pose as a viable solution for pod phenotyping, with potential applications in soybean breeding optimization.
Why it matches plant phenotyping methods大豆莢の表現型を抽出する深層学習キーポイント検出モデルを開発し、精度評価と複数形質の定量化を行っており、植物フェノタイピング手法が研究の中心である。
abstractWe propose Pod-pose, an innovative top-down keypoint detection model for precise soybean pod phenotyping
NeRF / 3D Gaussian SplattingLiDAR / point cloudWhole plant / canopy / plot / fieldPose / keypoint estimation2D/3D reconstructionSegmentationGrowth / development / phenology
Plant 3D reconstruction plays a critical role in precision agriculture and plant growth monitoring, yet it faces challenges such as complex background interference, difficulties in capturing intricate plant structures, and a slow reconstruction speed. In this study, we propose PlantSamGaussianReconstruction (PSGR), a novel method that integrates Grounding SAM with 3D Gaussian Splatting (3DGS) techniques. PSGR employs Grounding DINO and SAM for accurate plant–background segmentation, utilizes algorithms such as Scale-Invariant Feature Transform (SIFT) for camera pose estimation and sparse point cloud generation, and leverages 3DGS for plant reconstruction. Furthermore, a 3D–2D projection-guided optimization strategy is introduced to enhance segmentation precision. The experimental results of various multi-view plant image datasets demonstrate that PSGR effectively removes background noise under diverse environments, accurately captures plant details, and achieves peak signal-to-noise ratio (PSNR) values exceeding 30 in most scenarios, outperforming the original 3DGS approach. Moreover, PSGR reduces training time by up to 26.9%, significantly improving reconstruction efficiency. These results suggest that PSGR is an efficient, scalable, and high-precision solution for plant modeling.
Why it matches plant phenotyping methods植物のマルチビュー画像から3D構造を再構成する手法を開発・評価しており、植物形態の取得が研究の中心です。
abstractPlant 3D reconstruction plays a critical role in precision agriculture and plant growth monitoring
To obtain the complete shape and pose of corn under occlusion, this study proposes a point cloud completion algorithm for completing the fragmented corn point cloud after segmentation. Considering that this work focuses on a single-class crop—corn—the proposals mainly focus on the deep learning model size and the completion of the overall shape of the corn. In this work, the 3D corn models derived from segmentation are employed to systematically output the fragmented point cloud data in batches. The Shape Coding PointAttN (SCPAN) algorithm is also proposed, which is based on PointAttN. The model’s structure is simplified to output sparse point clouds and minimize computational complexity, and a gated multilayer perceptron (MLP) containing 3D position coding is introduced to enhance the model’s spatial awareness. In addition, the prior shape encoder module is initially trained and subsequently integrated into the model to enhance its focus on shape characteristics. Compared to the original model, PointAttN, SCPAN achieves a 34.2% reduction in the number of parameters, and the inference time is reduced by 30 ms while maintaining comparable accuracy. The experimental results show that the proposed method can complete the corn point cloud more effectively, using a small model to help estimate the pose and dimensions of corn accurately. This work supports the precise phenotypic analysis of corn and similar crops, such as citrus and tomatoes, and promotes the development of smart agricultural technology.
Why it matches plant phenotyping methodsトウモロコシの遮蔽点群を補完し、形状・姿勢・寸法を推定する計算手法の開発が中心であり、植物表現型取得への応用も明示されている。
abstractTo obtain the complete shape and pose of corn under occlusion, this study proposes a point cloud completion algorithm
Reproduction assets foundThe paper's Data Availability Statement points to an authors' GitHub repository for the corn point cloud completion code (SCPAN). The phenotype dataset itself is only available upon request from the corresponding author, so it does not qualify as a public asset.Code · publiccorresponding author upon reasonable request. The related code will be released at https://github.Open asset ↗pdf-page:14 lines:1-60Plant phenotyping relevance match · UnverifiedOpenAlex · checked 13 Sept 2026
To address the challenge of precise picking point localization in morphologically diverse safflower plants, this study proposes PointSafNet—a novel three-stage 3D point cloud analysis framework with distinct architectural and methodological innovations. In Stage I, we introduce a multi-view reconstruction pipeline integrating Structure from Motion (SfM) and Multi-View Stereo (MVS) to generate high-fidelity 3D plant point clouds. Stage II develops a dual-branch architecture employing Star modules for multi-scale hierarchical geometric feature extraction at the organ level (filaments and frui balls), complemented by a Context-Anchored Attention (CAA) mechanism to capture long-range contextual information. This synergistic feature learning approach addresses morphological variations, achieving 86.83% segmentation accuracy (surpassing PointNet++ by 7.37%) and outperforming conventional point cloud models. Stage III proposes an optimized geometric analysis pipeline combining dual-centroid spatial vectorization with Oriented Bounding Box (OBB)-based proximity analysis, resolving picking coordinate localization across diverse plants with 90% positioning accuracy and 68.82% mean IoU (13.71% improvement). The experiments demonstrate that PointSafNet systematically integrates 3D reconstruction, hierarchical feature learning, and geometric reasoning to provide visual guidance for robotic harvesting systems in complex plant canopies. The framework’s dual emphasis on architectural innovation and geometric modeling offers a generalizable solution for precision agriculture tasks involving morphologically diverse safflowers.
Why it matches plant phenotyping methods3D画像再構成・点群解析・器官セグメンテーションを統合し、植物器官の位置を推定する方法が研究の中心である。
abstractthis study proposes PointSafNet—a novel three-stage 3D point cloud analysis framework
Reliable, quantitative information on the presence and severity of crop diseases is essential for site-specific crop management and resistance breeding. Successful analysis of leaves under naturally variable lighting, presenting multiple disorders, and across phenological stages is a critical step towards high-throughput disease assessments directly in the field. Here, we present a dataset comprising 422 high resolution images of flattened leaves captured under variable outdoor lighting with polygon annotations of leaves, leaf necrosis and insect damage as well as point annotations of Septoria tritici blotch (STB) fruiting bodies (pycnidia) and rust pustules. Based on this dataset, we demonstrate the capability of deep learning for keypoint detection of pycnidia (F1=0.76) and rust pustules (F1=0.77) combined with semantic segmentation of leaves (IoU=0.96), leaf necrosis (IoU=0.77) and insect damage (IoU=0.69) to reliably detect and quantify the presence of STB, leaf rusts, and insect damage on symptom level under natural outdoor conditions. An analysis of intra- and inter-annotator agreement on selected images demonstrated that the proposed method achieved a performance close to that of annotators in the majority of the scenarios. We validated the generalization capabilities of the proposed method by testing it on images of unstructured canopies acquired directly in the field and without manual interaction with single leaves. This enables significantly higher throughput and automated data acquisition, which is critical to harness the full potential of image-based disease assessments. Model predictions were in good agreement with visual assessments of in-focus regions in these images, despite the presence of new challenges such as variable orientation of leaves and more complex lighting. This underscores the principle feasibility of diagnosing and quantifying the severity of foliar diseases under field conditions using the proposed imaging setup and image processing methods. By demonstrating the ability to diagnose and quantify the severity of multiple diseases in highly complex field scenarios, we lay the groundwork for high-throughput in-field assessments of foliar diseases that can support resistance breeding and the implementation of core principles of precision agriculture.
Why it matches plant phenotyping methods圃場画像から葉の壊死、病斑・病原体構造、害虫被害を検出・定量し、深層学習手法をデータセットで検証した研究であり、植物病害状態の表現型取得が中心です。
abstractHere, we present a dataset comprising 422 high resolution images of flattened leaves captured under variable outdoor lighting with polygon annotations of leaves, leaf necrosis and insect damage as well as point annotations of Septoria tritici blotch (STB) fruiting bodies (pycnidia) and rust pustules.
Posture is a critical phenotypic trait that reflects crop growth and serves as an essential indicator for both agricultural production and scientific research. Accurate pose estimation enables real-time tracking of crop growth processes, but in field environments, challenges such as variable backgrounds, dense planting, occlusions, and morphological changes hinder precise posture analysis. To address these challenges, we propose PFLO (Pose Estimation Model of Field Maize Based on YOLO Architecture), an end-to-end model for maize pose estimation, coupled with a novel data processing method to generate bounding boxes and pose skeleton data from a"keypoint-line"annotated phenotypic database which could mitigate the effects of uneven manual annotations and biases. PFLO also incorporates advanced architectural enhancements to optimize feature extraction and selection, enabling robust performance in complex conditions such as dense arrangements and severe occlusions. On a fivefold validation set of 1,862 images, PFLO achieved 72.2% pose estimation mean average precision (mAP50) and 91.6% object detection mean average precision (mAP50), outperforming current state-of-the-art models. The model demonstrates improved detection of occluded, edge, and small targets, accurately reconstructing skeletal poses of maize crops. PFLO provides a powerful tool for real-time phenotypic analysis, advancing automated crop monitoring in precision agriculture.
Why it matches plant phenotyping methodsトウモロコシの姿勢という植物表現型を、圃場画像から推定するモデルとデータ処理法を開発・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。
abstractwe propose PFLO (Pose Estimation Model of Field Maize Based on YOLO Architecture), an end-to-end model for maize pose estimation, coupled with a novel data processing method to generate bounding boxes and pose skeleton data
Reproduction assets found保存済みの本文根拠を更新済みルールで再検証し、公開資産1件を確認しました。Dataset · publicMIPDB dataset can be accessed at: http://phenomics.agis.org.cn/#/category. 2024;219:108795.Open asset ↗MIPDBpdf-page:25 lines:1-56Plant phenotyping relevance match · UnverifiedOpenAlex · checked 13 Sept 2026
Intelligent pruning technology is significant in reducing management costs and improving operational efficiency. In this study, a branch recognition and pruning point localization method was proposed for dormant walnut (Juglans regia L.) trees. First, 3D point clouds of walnut trees were reconstructed from multi-view images using Neural Radiance Fields (NeRFs). Second, Walnut-PointNet was improved to segment the walnut tree into Trunk, Branch, and Calibration categories. Next, individual pruning branches were extracted by cluster analysis and pruning rules were adjusted by classifying branches based on length. Finally, Principal Component Analysis (PCA) was used for length extraction, and pruning points were determined based on pruning rules. Walnut-PointNet achieved an OA of 93.39%, an ACC of 95.29%, and an mIoU of 0.912 on the walnut tree dataset. The mean absolute errors in length extraction for the short-growing branch group and the water sprout were 28.04 mm and 50.11 mm, respectively. The average success rate of pruning point recognition reached 89.33%, and the total time for pruning branch recognition and pruning point localization for the entire tree was approximately 16 s. This study provides support for the development of intelligent pruning for walnut trees.
Why it matches plant phenotyping methods樹木の3D点群から枝を認識・分割し、枝長を抽出する画像解析手法を開発・検証しており、植物器官形態の定量化が中心です。剪定点の局在化を含みますが、単なる対象検出にとどまらず枝長という再利用可能な植物形質を推定しています。
abstracta branch recognition and pruning point localization method was proposed for dormant walnut (Juglans regia L.) trees.
Corn is an important staple crop. Obtaining the pose and dimensions of corn is key for automating corn cultivation, mainly using robotic arms or similar devices to perform precise operations, such as pesticide spraying, measurement, or precise picking. In this study, the Stereo-Corn-Pose Detection (SCPD) algorithm was proposed, which used three dimensional (3D) object detection to obtain the pose and dimensions of corn with stereo images. This algorithm includes pitch angle detection, which is absent in traditional 3D object detection. The SCPD algorithm consists of two models: the union box detection model, FCOS-Stereo, based on the anchor-free network FCOSNet, and the 3D bounding box regression model, Cross-Stereo-EfficientFormer. This regression model incorporates a cross-attention mechanism into EfficientFormer to extract and fuse features effectively. This work constructed a dataset comprising 2,700 samples for training and 300 samples for testing. In the test set, this work achieved a union bounding box mAP of 85.3%, representing a 3.5% improvement over the original FCOSNet model. It also achieved 88.3% AP3D and 85.6% APBEV for 3D bounding box regression, making increases of 5% and 4.7%, respectively, compared to the traditional 3D object detection method, IDA-3D. The results indicate an accuracy of approximately 91% in detecting corn dimensions and pose. Therefore, the SCPD algorithm offers a novel framework for obtaining the 3D dimensions and pose of corn and promotes precision and smart agriculture for corn cultivation.
Why it matches plant phenotyping methodsトウモロコシの姿勢・寸法という植物形態形質をステレオ画像から推定する3D手法を開発し、データセットと精度評価も行っており、フェノタイピング手法が中心である。
abstractthe Stereo-Corn-Pose Detection (SCPD) algorithm was proposed, which used three dimensional (3D) object detection to obtain the pose and dimensions of corn with stereo images.
This paper presents a NeRF-based framework for point cloud (PCD) reconstruction, specifically designed for indoor high-throughput plant phenotyping facilities. Traditional NeRF-based reconstruction methods require cameras to move around stationary objects, but this approach is impractical for high-throughput environments where objects are rapidly imaged while moving on conveyors or rotating pedestals. To address this limitation, we develop a variant of NeRF-based PCD reconstruction that uses a single stationary camera to capture images as the object rotates on a pedestal. Our workflow comprises COLMAP-based pose estimation, a straightforward pose transformation to simulate camera movement, and subsequent standard NeRF training. A defined Region of Interest (ROI) excludes irrelevant scene data, enabling the generation of high-resolution point clouds (10M points). Experimental results demonstrate excellent reconstruction fidelity, with precision-recall analyses yielding an F-score close to 100.00 across all evaluated plant objects. Although pose estimation remains computationally intensive with a stationary camera setup, overall training and reconstruction times are competitive, validating the method's feasibility for practical high-throughput indoor phenotyping applications. Our findings indicate that high-quality NeRF-based 3D reconstructions are achievable using a stationary camera, eliminating the need for complex camera motion or costly imaging equipment. This approach is especially beneficial when employing expensive and delicate instruments, such as hyperspectral cameras, for 3D plant phenotyping. Future work will focus on optimizing pose estimation techniques and further streamlining the methodology to facilitate seamless integration into automated, high-throughput 3D phenotyping pipelines.
Why it matches plant phenotyping methods植物フェノタイピング施設向けに、固定カメラ画像からNeRFで植物の3D点群を再構成する手法を開発・評価しており、表現型取得が研究の中心である。
abstractThis paper presents a NeRF-based framework for point cloud (PCD) reconstruction, specifically designed for indoor high-throughput plant phenotyping facilities.
Reproduction assets foundThe paper explicitly releases its full SC-NeRF dataset (raw 4K videos, frames, COLMAP poses, NeRF checkpoints, and final 10M-point clouds for six plant/produce objects) on Hugging Face, and states that all datasets and the authors' code are available at the project page. Both are paper-specific, public, and actionable.Code · publicd delicate instruments, such as hyperspectral cameras, for 3D plant phenotyping. Future work will focus on optimizing pose estimation techniques and further streamlining the methodology to facilitate seamless integration into automated, high-throughput 3D phenotyping pipelines. We provide all datasets and our code, available at https://baskargroup.github.io/SC-NeRF/
Figure 1 : Schematic of the stationary camera imaging system for NeRF-based point cloud reconstruction in high-throughput plant phenotyping. In this setup, each plant is conveyed to a rotating turntable marked against a matte black background. Over a full 30-second rotation, a tripod-mounted stationary camera captures high-resoOpen asset ↗lines:1-53Plant phenotyping relevance match · UnverifiedOpenAlex · checked 14 Sept 2026
Leaf shape is of great significance in plant phenotype research. Landmarks method is a widely used morphometric approach, which can comprehensively describe the morphological differences among leaves. However, the selection of landmarks is time-consuming and laborious. An automatic landmarking algorithm is proposed here. Based on conformal mapping, the leaf outline can be transformed into a monotonically increasing function curve, referred to as the ’fingerprint function’. The Dynamic Time Warping (DTW) algorithm was introduced to match landmarks between different leaves. Two leaf datasets were used to validate the algorithm separately in different species and developmental stages. Dataset1 is a public dataset which covers 26 different types of leaves. The average positional difference between automatic and manual landmarks for dataset1 was only 2.95%. Dataset2 consists of cotton leaves collected in the field at various growth stages, and the positional difference for this dataset was all below 5%. These results validate that our algorithm is applicable to a wide range of leaf types and capable of identifying and locating novel features that emerge during leaf growth. The automatic landmarking algorithm can simulate manual landmarking to a great extent. It provides a new approach for automated acquisition of plant leaf shape homology tailored to the research needs of botanists.
Why it matches plant phenotyping methods葉形態の自動ランドマーク取得アルゴリズムを開発し、複数の葉データセットで手動測定と比較検証しているため、植物表現型の取得手法が中心である。
abstractAn automatic landmarking algorithm is proposed here.
NeRF / 3D Gaussian SplattingFlowerPose / keypoint estimation
This study presents Flower Pose Estimation (FloPE), a real-time flower pose estimation framework for computationally constrained robotic pollination systems. Robotic pollination has been proposed to supplement natural pollination to ensure global food security due to the decreased population of natural pollinators. However, flower pose estimation for pollination is challenging due to natural variability, flower clusters, and high accuracy demands due to the flowers' fragility when pollinating. This method leverages 3D Gaussian Splatting to generate photorealistic synthetic datasets with precise pose annotations, enabling effective knowledge distillation from a high-capacity teacher model to a lightweight student model for efficient inference. The approach was evaluated on both single and multi-arm robotic platforms, achieving a mean pose estimation error of 0.6 cm and 19.14 degrees within a low computational cost. Our experiments validate the effectiveness of FloPE, achieving up to 78.75% pollination success rate and outperforming prior robotic pollination techniques.
Why it matches plant phenotyping methods花の姿勢という植物器官の形態的状態を推定する画像・計算手法を開発し、ロボット実機で性能検証しており、フェノタイピング手法が中心である。
abstractThis study presents Flower Pose Estimation (FloPE), a real-time flower pose estimation framework for computationally constrained robotic pollination systems.
In agricultural practices, plant phenotyping using object detection models is gaining attention, plant phenotyping is a technology that accurately measures the quality and condition of cultivated crops from images, contributing to the improvement of crop yield and quality, as well as reducing environmental impact. However, collecting the training data necessary to create generic and high-precision models is extremely challenging due difficulties associated with annotations and the diversity of domains. Such difficulties arise from the unique shapes and backgrounds of plants, as well as the significant changes in appearance due to environmental conditions and growth stages. Furthermore, it is difficult to transfer training data across different crops, and although machine learning models effective for specific environments, conditions, and crops have been developed, they cannot be widely applied in real-world conditions. Therefore, in this study, we propose a generative artificial intelligence data augmentation method (D4) and investigated its application towards a shoot detection task in a vineyard. D4 uses a pre-trained text-guided diffusion model based on a large number of original images culled from video data collected by unmanned ground vehicles or other means, and a small number of annotated datasets. The proposed method generates new annotated images with background information adapted to the target domain while retaining annotation information necessary for object detection. In addition, D4 overcomes the lack of training data in agriculture, including the difficulty of annotation and diversity of domains. We confirmed that this generative data augmentation method improved the mean average precision by up to 28.65% for the BBox detection task and the average precision by up to 13.73% for the keypoint detection task for vineyard shoot detection. D4 generative data augmentation is expected to simultaneously solve the cost and domain diversity issues of training data generation for agricultural applications and improve the generalization performance of detection models.
Why it matches plant phenotyping methodsブドウのシュート検出を対象に、テキスト誘導拡散モデルによるドメイン適応型データ拡張手法を開発・評価しており、画像から植物器官を抽出する方法が研究の中心である。
abstractTherefore, in this study, we propose a generative artificial intelligence data augmentation method (D4) and investigated its application towards a shoot detection task in a vineyard.
Spectral imaging has been widely applied for soybean phenotyping to find and maintain favorable traits. Specifically, in soybean phenotyping, hyperspectral imaging through contact-based proximal sensing demonstrates better signal-to-noise ratio and resolution compared to remote sensing. However, it has not been adapted for large-scale field applications due to its low throughput and high labor costs. Additionally, no automation solution has been developed to collect in vivo contact-based hyperspectral images of soybean plants. In this study, a novel drone-based robotic system was developed to automate the collection of in vivo contact-based hyperspectral images in the field. The system consists of a machine vision system to detect and estimate the pose of soybean leaflets, an articulated robotic arm with specialized control and path planning algorithms to operate contact-based sensors to grasp and image the leaf, and a customized high-payload drone to provide mobility for sampling at different locations across a field. The average accuracy of the optimized machine vision algorithm is 95.88% for leaf detection and 97.54% for leaf pose estimation, and the average success rate of leaf grasping is 90.55%. This study presents an innovative method for expanding the applicability in vivo contact-based hyperspectral imaging for extensive agricultural applications.
Why it matches plant phenotyping methods植物葉の接触型ハイパースペクトル画像を自動取得するロボットシステムと、葉検出・姿勢推定・把持の技術を開発しており、植物フェノタイピング手法が研究の中心である。
abstractIn this study, a novel drone-based robotic system was developed to automate the collection of in vivo contact-based hyperspectral images in the field.
TomatoLiDAR / point cloudFruitObject detectionPose / keypoint estimationSegmentation
Tomatoes are widely grown all over the world and are one of the favourite vegetables of humanity. Tomato harvesting requires a lot of labor. With the aging population and the increasing demand for tomatoes, it is imminent to develop tomato picking robots. In the picking environment with a complex background, a 3D pose of the target is essential. It can provide guidance for robotic arm attitude and obstacle avoidance during picking. In this paper, a tomato pose detection algorithm (TPD) is proposed for 3D pose detection of a single fruit of clustered tomato. The TPD algorithm is divided into two modules: YOLO-lmk model and the point cloud processing module. By analyzing the network structure, optimizing parameters, optimizing loss function, adding attention mechanism and adding keypoint prediction, a YOLO-lmk model with YOLO v5s as the core is proposed to realize tomato bounding box and keypoint detection. The point cloud processing module is composed of point cloud segmentation, voxel downsampling, removing outliers, Euclidean color clustering, RANSAC sphere fitting, and keypoint index. As the experimental results show, YOLO-lmk model bounding box mAP is 92.9%, dₗₘₖ is 7.9, FLoating point Operations Per Second (FLOPs) is 16.6B, and speed is 0.062s/sheet. The dₗₘₖ represents the Euclidean distance between true keypoints and predicted keypoints. Compared with YOLO v5s, mAP is increased 1.8%, and FLOPs is increased only 0.1B. The point cloud processing module only takes 0.028s to complete a tomato 3D pose detection, which is fast. The accuracy of the TPD algorithm in detecting tomatoes is 93.4%, and the time cost is 0.09 s for detecting one tomato. The TPD algorithm can provide a theoretical basis for tomato 3D pose detection, and provide a reference for other fruits (pears, citrus, apples) 3D pose detection.
Why it matches plant phenotyping methodsトマト果実の3D姿勢という植物器官の形態状態を、画像・点群処理で推定する手法を開発・評価しており、収穫ロボット向けでも再利用可能な表現型抽出が中心である。
abstractIn this paper, a tomato pose detection algorithm (TPD) is proposed for 3D pose detection of a single fruit of clustered tomato.
CitrusField / plotLiDAR / point cloudFruitObject detectionPose / keypoint estimation
The growth poses of citrus on trees are diverse. To ensure minimal loss during citrus harvesting, accurately estimating the pose of citrus is particularly important. To solve this problem, this research developed a real-time citrus pose estimation system based on neural networks and point cloud processing algorithms. Specifically, this method uses neural networks to identify citrus. After constructing the citrus point cloud, it is input into the Random Sample Consensus with Levenberg-Marquardt (RANSAC-LM) point cloud processing algorithm to obtain the citrus coordinates. Combined with citrus growth information, the pose is output. By analyzing the distribution of citrus poses, citrus poses convenient for end- effector harvesting are defined. To enhance the camera's ability to obtain information about citrus, a camera observation model is constructed to dynamically adjust the camera position. Through experiments, the appropriate deep learning target detection framework YOLO V5 is selected for citrus object detection. The precision (P), recall rate (R), and mean average precision (mAP) are 92.3 %, 79.1 %, and 88.5 % respectively. This network can handle detection tasks in real orchard environments. The original Random Sample Consensus (RANSAC) is improved by using the Levenberg-Marquardt (LM) nonlinear optimization method. Experimental results show that RANSAC-LM reduces the citrus center coordinate precision error from (0.2, 0.2, 2.3) mm to (0.1, 0.2, 1.4) mm, reduces the accuracy Spherical Error Probable (SEP) from 2.77 to 1.61, and finally reduces the citrus pose error from 5.72° to 2.43°. The efficiency of the proposed citrus pose estimation algorithm is 0.24 s. Deployed on a citrus picking robot, it verifies the feasibility of the algorithm and provides a new solution for the pose estimation problem of citrus harvesting robots.
Why it matches plant phenotyping methods柑橘果実の座標・生育姿勢という器官形質を点群処理とニューラルネットワークで推定し、精度を実験検証している。収穫対象の単なる検出を超えた姿勢計測法が中心である。
abstractthis research developed a real-time citrus pose estimation system based on neural networks and point cloud processing algorithms
In agricultural practices, plant phenotyping using object detection models is gaining attention, plant phenotyping is a technology that accurately measures the quality and condition of cultivated crops from images, contributing to the improvement of crop yield and quality, as well as reducing environmental impact. However, collecting the training data necessary to create generic and high-precision models is extremely challenging due difficulties associated with annotations and the diversity of domains. Such difficulties arise from the unique shapes and backgrounds of plants, as well as the significant changes in appearance due to environmental conditions and growth stages. Furthermore, it is difficult to transfer training data across different crops, and although machine learning models effective for specific environments, conditions, and crops have been developed, they cannot be widely applied in real-world conditions. Therefore, in this study, we propose a generative artificial intelligence data augmentation method (D4) and investigated its application towards a shoot detection task in a vineyard. D4 uses a pre-trained text-guided diffusion model based on a large number of original images culled from video data collected by unmanned ground vehicles or other means, and a small number of annotated datasets. The proposed method generates new annotated images with background information adapted to the target domain while retaining annotation information necessary for object detection. In addition, D4 overcomes the lack of training data in agriculture, including the difficulty of annotation and diversity of domains. We confirmed that this generative data augmentation method improved the mean average precision by up to 28.65% for the BBox detection task and the average precision by up to 13.73% for the keypoint detection task for vineyard shoot detection. D4 generative data augmentation is expected to simultaneously solve the cost and domain diversity issues of training data generation for agricultural applications and improve the generalization performance of detection models. • Proposed novel data augmentation method D4 using text-guided diffusion model. • Analyzed detection accuracy using D4 for BBox detection and keypoint detection. • D4 improved detection accuracy and demonstrated effectiveness of domain adaptation. • D4 uses automatic selection with DreamSim to maintain generated image quality. • D4 overcomes lack of training data in agriculture.
Why it matches plant phenotyping methodsブドウのシュートを画像から検出する植物フェノタイピング向けのデータ拡張・ドメイン適応手法を開発し、検出精度を検証しており、表現型取得手法が中心である。
abstractTherefore, in this study, we propose a generative artificial intelligence data augmentation method (D4) and investigated its application towards a shoot detection task in a vineyard.
Hyperspectral imaging (HSI) is a prevalent method in crop phenotyping. Nevertheless, current HSI remote sensing techniques are compromised by changing ambient lighting conditions, long imaging distances, and comparatively low resolutions. Proximal HSI sensors such as LeafSpec were developed to improve the imaging quality. However, the application of proximal sensors remains contrained by their low throughput and intensive labor costs. Moreover, few automation solutions were available to use LeafSpec in phenotyping dicot plants. In this paper, a novel robotic system is presented as a sensor platform to operate LeafSpec to collect leaf-level hyperspectral images for in vivo phenotyping of soybean. A machine vision algorithm was developed to detect the top mature trifoliate and estimate the poses of the leaflets. A control and motion planning algorithm was developed for an articulated robotic manipulator to grasp the target leaflets. An experiment was conducted in March 2021 in a greenhouse with 64 soybean plants of 2 genotypes and 2 nitrogen treatments. The machine vision detected the target leaflets with a first trial success rate of 84.13% and an overall success rate of 90.66%. The robotic manipulator operated LeafSpec to image the target leaflets with a first trial success rate of 87.30% and an overall success rate of 93.65%. The average cycle time for one soybean plant was 63.20 s. The PLS predictions from the robot-collected data had an R² of 0.84 with the measured nitrogen content and an R² of 0.82 with the predictions from human-collected data. The results demonstrated the potential of applying the system for automated in vivo leaf-level HSI for soybean phenotyping in the field.
Why it matches plant phenotyping methodsロボットによる近接ハイパースペクトル画像取得システムと葉検出・動作計画アルゴリズムを開発し、豆類の表現型計測性能を検証しており、方法が中心的である。
abstractIn this paper, a novel robotic system is presented as a sensor platform to operate LeafSpec to collect leaf-level hyperspectral images for in vivo phenotyping of soybean.
Phenotypic traits, such as plant height, internode length, and node count, are essential indicators of the growth status of tomato plants, carrying significant implications for research on genetic breeding and cultivation management. Deep learning algorithms such as object detection and segmentation have been widely utilized to extract plant phenotypic parameters. However, segmentation-based methods are labor-intensive due to their requirement for extensive annotation during training, while object detection approaches exhibit limitations in capturing intricate structural features. To achieve real-time, efficient, and precise extraction of phenotypic traits of seedling tomatoes, a novel plant phenotyping approach based on 2D pose estimation was proposed. We enhanced a novel heatmap-free method, YOLOv8s-pose, by integrating the Convolutional Block Attention Module (CBAM) and Content-Aware ReAssembly of FEatures (CARAFE), to develop an improved YOLOv8s-pose (IYOLOv8s-pose) model, which efficiently focuses on salient image features with minimal parameter overhead while achieving a superior recognition performance in complex backgrounds. IYOLOv8s-pose manifested a considerable enhancement in detecting bending points and stem nodes. Particularly for internode detection, IYOLOv8s-pose attained a Precision of 99.8%, exhibiting a significant improvement over RTMPose-s, YOLOv5s6-pose, YOLOv7s-pose, and YOLOv8s-pose by 2.9%, 5.4%, 3.5%, and 5.4%, respectively. Regarding plant height estimation, IYOLOv8s-pose achieved an RMSE of 0.48 cm and an rRMSE of 2%, and manifested a 65.1%, 68.1%, 65.6%, and 51.1% reduction in the rRMSE compared to RTMPose-s, YOLOv5s6-pose, YOLOv7s-pose, and YOLOv8s-pose, respectively. When confronted with the more intricate extraction of internode length, IYOLOv8s-pose also exhibited a 15.5%, 23.9%, 27.2%, and 12.5% reduction in the rRMSE compared to RTMPose-s, YOLOv5s6-pose, YOLOv7s-pose, and YOLOv8s-pose. IYOLOv8s-pose achieves high precision while simultaneously enhancing efficiency and convenience, rendering it particularly well suited for extracting phenotypic parameters of tomato plants grown naturally within greenhouse environments. This innovative approach provides a new means for the rapid, intelligent, and real-time acquisition of plant phenotypic parameters in complex backgrounds.
Why it matches plant phenotyping methodsトマトの草丈・節間長・節数を画像から抽出する2Dポーズ推定法を開発し、既存手法と性能比較・検証しており、植物表現型取得が研究の中心である。
abstractTo achieve real-time, efficient, and precise extraction of phenotypic traits of seedling tomatoes, a novel plant phenotyping approach based on 2D pose estimation was proposed.
• UAV-based localization and mapping methods have been benchmarked in vineyards. • Five evaluation metrics were developed for agricultural scenarios. • Lighting variation impacts point cloud resolution. • Deep learning enhances SLAM for efficient plant phenotyping. UAVs equipped with various sensors offer a promising approach for enhancing orchard management efficiency. Up-close sensing enables precise crop localization and mapping, providing valuable a priori information for informed decision-making. Current research on localization and mapping methods can be broadly classified into SfM, traditional feature-based SLAM, and deep learning-integrated SLAM. While previous studies have evaluated these methods on public datasets, real-world agricultural environments, particularly vineyards, present unique challenges due to their complexity, dynamism, and unstructured nature. To bridge this gap, we conducted a comprehensive study in vineyards, collecting data under diverse conditions (flight modes, illumination conditions, and shooting angles) using a UAV equipped with high-resolution camera. To assess the performance of different methods, we proposed five evaluation metrics: efficiency, point cloud completeness, localization accuracy, parameter sensitivity, and plant-level spatial accuracy. We compared two SLAM approaches against SfM as a benchmark. Our findings reveal that deep learning-based SLAM outperforms SfM and feature-based SLAM in terms of position accuracy and point cloud resolution. Deep learning-based SLAM reduced average position error by 87% and increased point cloud resolution by 571%. However, feature-based SLAM demonstrated superior efficiency, making it a more suitable choice for real-time applications. These results offer valuable insights for selecting appropriate methods, considering illumination conditions, and optimizing parameters to balance accuracy and computational efficiency in orchard management activities.
Why it matches plant phenotyping methodsブドウ園でのUAV画像によるSfM・SLAM手法を比較検証し、植物レベルの空間精度や点群完全性などを評価しており、植物フェノタイピングの取得・解析基盤が中心である。
abstractDeep learning enhances SLAM for efficient plant phenotyping.
Plant phenotyping relevance match · UnverifiedOpenAlex · Crossref · Europe PMC · checked 7 Sept 2026
Cucumis melo L., commonly known as melon, is a crucial horticultural crop. The selection and breeding of superior melon germplasm resources play a pivotal role in enhancing its marketability. However, current methods for melon appearance phenotypic analysis rely primarily on expert judgment and intricate manual measurements, which are not only inefficient but also costly. Therefore, to expedite the breeding process of melon, we analyzed the images of 117 melon varieties from two annual years utilizing artificial intelligence (AI) technology. By integrating the semantic segmentation model Dual Attention Network (DANet), the object detection model RTMDet, the keypoint detection model RTMPose, and the Mobile-Friendly Segment Anything Model (MobileSAM), a deep learning algorithm framework was constructed, capable of efficiently and accurately segmenting melon fruit and pedicel. On this basis, a series of feature extraction algorithms were designed, successfully obtaining 11 phenotypic traits of melon. Linear fitting verification results of selected traits demonstrated a high correlation between the algorithm-predicted values and manually measured true values, thereby validating the feasibility and accuracy of the algorithm. Moreover, cluster analysis using all traits revealed a high consistency between the classification results and genotypes. Finally, a user-friendly software was developed to achieve rapid and automatic acquisition of melon phenotypes, providing an efficient and robust tool for melon breeding, as well as facilitating in-depth research into the correlation between melon genotypes and phenotypes.
Why it matches plant phenotyping methodsメロン果実・花柄画像から11形質を抽出する深層学習・画像解析フレームワークを開発し、手動測定との検証とソフトウェア化まで行っており、植物表現型取得法が中心である。
abstracta deep learning algorithm framework was constructed, capable of efficiently and accurately segmenting melon fruit and pedicel.
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
Accurately detecting the maturity and 3D position of flowering Chinese cabbage ( Brassica rapa var. chinensis) in natural environments is vital for autonomous robot harvesting in unstructured farms. The challenge lies in dense planting, small flower buds, similar colors and occlusions. This study proposes a YOLOv8-Improved network integrated with the ByteTrack tracking algorithm to achieve multi-object detection and 3D positioning of flowering Chinese cabbage plants in fields. In this study, C2F-MLCA is created by adding a lightweight Mixed Local Channel Attention (MLCA) with spatial awareness capability to the C2F module of YOLOv8, which improves the extraction of spatial feature information in the backbone network. In addition, a P2 detection layer is added to the neck network, and BiFPN is used instead of PAN to enhance multi-scale feature fusion and small target detection. Wise-IoU in combination with Inner-IoU is adopted as a new loss function to optimize the network for different quality samples and different size bounding boxes. Lastly, ByteTrack is integrated for video tracking, and RGB-D camera depth data are used to estimate cabbage positions. The experimental results show that YOLOv8-Improve achieves a precision ( P ) of 86.5% and a recall ( R ) of 86.0% in detecting the maturity of flowering Chinese cabbage. Among them, mAP50 and mAP75 reach 91.8% and 61.6%, respectively, representing an improvement of 2.9% and 4.7% over the original network. Additionally, the number of parameters is reduced by 25.43%. In summary, the improved YOLOv8 algorithm demonstrates high robustness and real-time detection performance, thereby providing strong technical support for automated harvesting management.
Why it matches plant phenotyping methods開花中国白菜の成熟度という植物状態を画像から検出・推定する改良YOLOv8と3D位置推定手法が研究の中心であり、収穫対象の単なる位置検出を超える。
abstractThis study proposes a YOLOv8-Improved network integrated with the ByteTrack tracking algorithm to achieve multi-object detection and 3D positioning of flowering Chinese cabbage plants in fields.
The estimation of tea leaf pose is an emerging research topic. Recognising the morphological features of tea leaves can help accurately categorise, grade, and determine their level of maturity. Therefore, this study proposes a deep neural network, TeaPoseNet, to estimate tea leaf poses. The algorithm was trained and validated using a dataset of one-bud-one-leaf images of Yinghong No.9 tea leaves and was compared with four other pose estimation networks. At the same time, the contribution of TKS_NMS to the algorithm was validated through ablation experiments. The results indicate that TKS_NMS improved the EPE accuracy of pose recognition by 16.33 %. More specifically, the algorithm achieved a good overall performance, with PCK, AUC, EPE, and NME reaching 0.9800, 0.8147, 9.0955, and 0.0644, respectively. The average running speed for detecting the pose of a single tea leaf image was 40.01 ms. To the best of our knowledge, this is the first application of pose estimation technology to the detection and analysis of Yinghong No.9 tea leaves. The results show that the proposed algorithm can effectively estimate the pose of tea leaves, thus providing a reference for subsequent tea research.
Why it matches plant phenotyping methods茶葉の葉姿勢という植物形態形質を画像から推定する深層学習手法を開発し、データセット、比較評価、アブレーション検証、性能指標を提示しており、植物フェノタイピング手法が研究の中心である。
abstractTherefore, this study proposes a deep neural network, TeaPoseNet, to estimate tea leaf poses.
Introduction Accurate and rapid identification of cabbage posture is crucial for minimizing damage to cabbage heads during mechanical harvesting. However, due to the structural complexity of cabbages, current methods encounter challenges in detecting and segmenting the heads and roots. Therefore, exploring efficient cabbage posture prediction methods is of great significance. Methods This study introduces YOLOv5-POS, an innovative cabbage posture prediction approach. Building on the YOLOv5s backbone, this method enhances detection and segmentation capabilities for cabbage heads and roots by incorporating C-RepGFPN to replace the traditional Neck layer, optimizing feature extraction and upsampling strategies, and refining the C-Seg segmentation head. Additionally, a cabbage root growth prediction model based on Bézier curves is proposed, using the geometric moment method for key point identification and the anti-gravity stem-seeking principle to determine root-head junctions. It performs precision root growth curve fitting and prediction, effectively overcoming the challenge posed by the outer leaves completely enclosing the cabbage root stem. Results and discussion YOLOv5-POS was tested on a multi-variety cabbage dataset, achieving an F1 score of 98.8% for head and root detection, with an instance segmentation accuracy of 93.5%. The posture recognition model demonstrated an average absolute error of 1.38° and an average relative error of 2.32%, while the root growth prediction model reached an accuracy of 98%. Cabbage posture recognition was completed within 28 milliseconds, enabling real-time harvesting. The enhanced model effectively addresses the challenges of cabbage segmentation and posture prediction, providing a highly accurate and efficient solution for automated harvesting, minimizing crop damage, and improving operational efficiency.
Why it matches plant phenotyping methodsキャベツの頭部・根の検出／セグメンテーションと姿勢角・根の成長曲線を画像から推定する手法を開発し、精度と処理時間を評価しており、植物形質取得が中心である。
abstractThis study introduces YOLOv5-POS, an innovative cabbage posture prediction approach.
In an agricultural field, plant phenotyping using object detection models is gaining attention. However, collecting the training data necessary to create generic and high-precision models is extremely challenging due to the difficulty of annotation and the diversity of domains. Furthermore, it is difficult to transfer training data across different crops, and although machine learning models effective for specific environments, conditions, or crops have been developed, they cannot be widely applied in actual fields. In this study, we propose a generative data augmentation method (D4) for vineyard shoot detection. D4 uses a pre-trained text-guided diffusion model based on a large number of original images culled from video data collected by unmanned ground vehicles or other means, and a small number of annotated datasets. The proposed method generates new annotated images with background information adapted to the target domain while retaining annotation information necessary for object detection. In addition, D4 overcomes the lack of training data in agriculture, including the difficulty of annotation and diversity of domains. We confirmed that this generative data augmentation method improved the mean average precision by up to 28.65% for the BBox detection task and the average precision by up to 13.73% for the keypoint detection task for vineyard shoot detection. Our generative data augmentation method D4 is expected to simultaneously solve the cost and domain diversity issues of training data generation in agriculture and improve the generalization performance of detection models.
Why it matches plant phenotyping methodsブドウのシュート検出を対象に、ドメイン適応型の生成データ拡張法を開発・評価しており、植物器官の画像ベース表現型取得における方法的貢献が中心である。
abstractIn this study, we propose a generative data augmentation method (D4) for vineyard shoot detection.
Plant phenotype detection plays a crucial role in understanding and studying plant biology, agriculture, and ecology. It involves the quantification and analysis of various physical traits and characteristics of plants, such as plant height, leaf shape, angle, number, and growth trajectory. By accurately detecting and measuring these phenotypic traits, researchers can gain insights into plant growth, development, stress tolerance, and the influence of environmental factors, which has important implications for crop breeding. Among these phenotypic characteristics, the number of leaves and growth trajectory of the plant are most accessible. Nonetheless, obtaining these phenotypes is labor intensive and financially demanding. With the rapid development of computer vision technology and artificial intelligence, using maize field images to fully analyze plant-related information can greatly eliminate repetitive labor and enhance the efficiency of plant breeding. However, it is still difficult to apply deep learning methods in field environments to determine the number and growth trajectory of leaves and stalks due to the complex backgrounds and serious occlusion problems of crops in field environments. To preliminarily explore the application of deep learning technology to the acquisition of the number of leaves and stalks and the tracking of growth trajectories in field agriculture, in this study, we developed a deep learning method called Point-Line Net, which is based on the Mask R-CNN framework, to automatically recognize maize field RGB images and determine the number and growth trajectory of leaves and stalks. The experimental results demonstrate that the object detection accuracy (mAP50) of our Point-Line Net can reach 81.5%. Moreover, to describe the position and growth of leaves and stalks, we introduced a new lightweight "keypoint" detection branch that achieved a magnitude of 33.5 using our custom distance verification index. Overall, these findings provide valuable insights for future field plant phenotype detection, particularly for datasets with dot and line annotations.
Why it matches plant phenotyping methodsトウモロコシの葉・茎の数と成長軌跡をRGB画像から抽出する深層学習手法を開発し、精度評価も行っており、植物表現型取得が研究の中心である。
abstractin this study, we developed a deep learning method called Point-Line Net, which is based on the Mask R-CNN framework, to automatically recognize maize field RGB images and determine the number and growth trajectory of leaves and stalks.
Reproduction assets foundThe authors explicitly deposit the code (and data) supporting this maize leaf/stalk trajectory phenotyping study in a public GitHub repository, matching an allowed URL.Code · publicThe computer code and data that support the findings of this study are deposited in a GitHub repository at https://github.com/VEGETALOADING/Point-Line-Net .Open asset ↗VEGETALOADING/Point-Line-Netlines:319-524Plant phenotyping relevance match · UnverifiedEurope PMC · checked 15 Sept 2026
Greenhouse production of fruits and vegetables in developed countries is challenged by labour scarcity and high labour costs. Robots offer a good solution for sustainable and cost-effective production. Acquiring accurate spatial information about relevant plant parts is vital for successful robot operation. Robot perception in greenhouses is challenging due to variations in plant appearance, viewpoints, and illumination. This paper proposes a keypoint-detection-based method using data from an RGB-D camera to estimate the 3D pose of peduncle nodes, which provides essential information to harvest the tomato bunches. Specifically, this paper proposes a method that detects four anatomical landmarks in the colour image and then integrates 3D point-cloud information to determine the 3D pose. A comprehensive evaluation was conducted in a commercial greenhouse to gain insight into the performance of different parts of the method. The results showed: (1) high accuracy in object detection, achieving an Average Precision (AP) of AP@0.5=0.96; (2) an average Percentage of Detected Joints (PDJ) of the keypoints of PhDJ@0.2 = 94.31%; and (3) 3D pose estimation accuracy with mean absolute errors (MAE) of 11ᵒ and 10ᵒ for the relative upper and lower angles between the peduncle and main stem, respectively. Furthermore, the capability to handle variations in viewpoint was investigated, demonstrating the method was robust to view changes. However, canonical and higher views resulted in slightly higher performance compared to other views. Although tomato was selected as a use case, the proposed method has the potential to be applied to other greenhouse crops, such as pepper, after fine-tuning.
Why it matches plant phenotyping methodsRGB-D画像と深度点群を用いてトマトの器官ランドマークを検出し、ペドンクル節の3D姿勢を推定する手法を開発・評価しており、植物形態の取得が中心である。
abstractThis paper proposes a keypoint-detection-based method using data from an RGB-D camera to estimate the 3D pose of peduncle nodes
The pod and seed counts are important yield-related traits in soybean. High-precision soybean breeders face the major challenge of accurately phenotyping the number of pods and seeds in a high-throughput manner. Recent advances in artificial intelligence, especially deep learning (DL) models, have provided new avenues for high-throughput phenotyping of crop traits with increased precision. However, the available DL models are less effective for phenotyping pods that are densely packed and overlap in in situ soybean plants; thus, accurate phenotyping of the number of pods and seeds in soybean plant is an important challenge. To address this challenge, the present study proposed a bottom-up model, DEKR-SPrior (disentangled keypoint regression with structural prior), for in situ soybean pod phenotyping, which considers soybean pods and seeds analogous to human people and joints, respectively. In particular, we designed a novel structural prior (SPrior) module that utilizes cosine similarity to improve feature discrimination, which is important for differentiating closely located seeds from highly similar seeds. To further enhance the accuracy of pod location, we cropped full-sized images into smaller and high-resolution subimages for analysis. The results on our image datasets revealed that DEKR-SPrior outperformed multiple bottom-up models, viz., Lightweight-OpenPose, OpenPose, HigherHRNet, and DEKR, reducing the mean absolute error from 25.81 (in the original DEKR) to 21.11 (in the DEKR-SPrior) in pod phenotyping. This paper demonstrated the great potential of DEKR-SPrior for plant phenotyping, and we hope that DEKR-SPrior will help future plant phenotyping.
Why it matches plant phenotyping methods大豆の莢・種子数を高スループットに推定する画像解析モデルを開発し、既存モデルと比較検証しているため、植物表現型取得手法が研究の中心です。
abstractthe present study proposed a bottom-up model, DEKR-SPrior (disentangled keypoint regression with structural prior), for in situ soybean pod phenotyping
Reproduction assets foundThe paper's DEKR-SPrior analysis code is publicly available on GitHub with an explicit availability statement and URL. The homemade soybean pod/seed image datasets are not publicly deposited and require contacting the corresponding author.Code · publicThe source code is publicly available. It can be accessed at the following GitHub repository: https://github.com/Cyncihe/DEKR-SPrior.gitOpen asset ↗https://github.com/Cyncihe/DEKR-SPrior.gitlines:300-403Plant phenotyping relevance match · UnverifiedEurope PMC · Crossref · OpenAlex · checked 7 Sept 2026
Abstract A method integrating camera pose recovery techniques with neural radiation field theory is proposed in this study to address issues such as detail loss and color distortion encountered by traditional stereoscopic vision-based 3D reconstruction techniques when dealing with fruit trees exhibiting high-frequency phenotypic details. The high cost of information acquisition devices equipped with image pose recording functionality necessitates a cost-effective approach for fruit tree information gathering while enhancing the resolution and detail capture capability of the resulting 3D models. To achieve this, a device and scheme for capturing multi-view image sequences of fruit trees are designed. Firstly, the target fruit tree is surrounded by a multi-angle video capture using the information acquisition platform, and the resulting video undergoes image enhancement and frame extraction to obtain a multi-view image sequence of the fruit tree. Subsequently, a motion recovery structure algorithm is employed for sparse reconstruction to recover image poses. Then, the image sequence with pose data is inputted into a multi-layer perceptron, utilizing ray casting for coarse and fine two-layer granularity sampling to calculate volume density and RGB information, thereby obtaining the neural radiation field 3D scene of the fruit tree. Finally, the 3D scene is converted into point clouds to derive a high-precision point cloud model of the fruit tree. Using this reconstruction method, a crabapple tree including multiple periods such as flowering, fruiting, leaf fall, and dormancy is reconstructed, capturing the neural radiation field scenes and point cloud models. Reconstruction results demonstrate that the 3D scenes of the neural radiation field in each period exhibit real-world level representation. The point cloud models derived from the 3D scenes achieve millimeter-level precision at the organ scale, with tree structure accuracy exceeding 96% for multi-period point cloud models, averaging 97.79% accuracy across all periods. This reconstruction method exhibits robustness across various fruit tree periods and can meet the requirements for 3D reconstruction of fruit trees in most scenarios.
Why it matches plant phenotyping methods果樹の多視点画像からNeRFと点群を用いて樹体・器官スケールの3D形態を再構成する手法を開発し、精度と複数生育期での頑健性を評価しているため、植物フェノタイピング手法が中心である。
abstractA method integrating camera pose recovery techniques with neural radiation field theory is proposed in this study to address issues such as detail loss and color distortion encountered by traditional stereoscopic vision-based 3D reconstruction techniques when dealing with fruit trees exhibiting high-frequency phenotypic details.
Introduction Yunnan Xiaomila is a pepper variety whose flowers and fruits become mature at the same time and multiple times a year. The distinction between the fruits and the background is low and the background is complex. The targets are small and difficult to identify. Methods This paper aims at the problem of target detection of Yunnan Xiaomila under complex background environment, in order to reduce the impact caused by the small color gradient changes between xiaomila and background and the unclear feature information, an improved PAE-YOLO model is proposed, which combines the EMA attention mechanism and DCNv3 deformable convolution is integrated into the YOLOv8 model, which improves the model's feature extraction capability and inference speed for Xiaomila in complex environments, and achieves a lightweight model. First, the EMA attention mechanism is combined with the C2f module in the YOLOv8 network. The C2f module can well extract local features from the input image, and the EMA attention mechanism can control the global relationship. The two complement each other, thereby enhancing the model's expression ability; Meanwhile, in the backbone network and head network, the DCNv3 convolution module is introduced, which can adaptively adjust the sampling position according to the input feature map, contributing to stronger feature capture capabilities for targets of different scales and a lightweight network. It also uses a depth camera to estimate the posture of Xiaomila, while analyzing and optimizing different occlusion situations. The effectiveness of the proposed method was verified through ablation experiments, model comparison experiments and attitude estimation experiments. Results The experimental results indicated that the model obtained an average mean accuracy (mAP) of 88.8%, which was 1.3% higher than that of the original model. Its F1 score reached 83.2, and the GFLOPs and model sizes were 7.6G and 5.7MB respectively. The F1 score ranked the best among several networks, with the model weight and gigabit floating-point operations per second (GFLOPs) being the smallest, which are 6.2% and 8.1% lower than the original model. The loss value was the lowest during training, and the convergence speed was the fastest. Meanwhile, the attitude estimation results of 102 targets showed that the orientation was correctly estimated exceed 85% of the cases, and the average error angle was 15.91°. In the occlusion condition, 86.3% of the attitude estimation error angles were less than 40°, and the average error angle was 23.19°. Discussion The results show that the improved detection model can accurately identify Xiaomila targets fruits, has higher model accuracy, less computational complexity, and can better estimate the target posture.
Why it matches plant phenotyping methods唐辛子果実の検出に加え、深度カメラによる姿勢推定手法を開発し、アブレーション・比較・姿勢推定実験で技術検証しているため、植物器官の表現型取得が中心である。
In facility horticulture smart farms, harvesting robotic systems to automate harvesting tasks are challenging due to the complex environment, irregular growth and fruits pose. To harvest a fruits, the vision information required by a harvesting robot is accurate position and pose. Especially, fruit pose information is essential for planning paths to avoid damaging stems, leaves, branches, and obstacles, preventing damage to the fruit and harvesting robots, and planning efficient harvest sequences. This paper presents a method that uses information from tomato sepals to estimate the orientation of tomatoes, which are a commonly grown crop in horticulture. First, we train a YOLOv8 model to detect and segment bodies and sepals of tomatoes. For robust training of the model, the training data is constructed using effective data augmentation, which synthesizes segmented foreground objects and inserting them into the background. Then, for finding association between the body and sepal of a tomato, we apply IoU score based matching and the Hungarian algorithm. Consequently, we obtain point clouds for both parts using RGB-D data. Finally, we compute the center point of each object by using spherical fitting and the statistics of the point cloud, respectively. Then, we estimate the pose of the tomato as a vector between two center points of the body and sepal. To accurately and practically evaluate the proposed method, we generated ground truth data using calibration patterns in tomato greenhouse. As the experimental results show, The segmentation results show that AP50,sepal is 94.7, AP50,tomato is 96.3, and mAP is 61.5. The result of the pose estimation for the pose validation dataset is a mean error angle of 6.79 ± 3.18 and angle errors of less than 10 degrees account for 87.2%. The total algorithm proposed in this paper requires about 0.038s for each tomato greenhouse image. As a result, our proposed pose estimation can be practically utilized in robot systems for tomato harvesting.
Why it matches plant phenotyping methodsトマト果実の姿勢という植物器官の形態特性を、画像・RGB-Dデータから推定する手法を開発し、精度検証まで行っているため、中心的な植物フェノタイピング研究である。
abstractThis paper presents a method that uses information from tomato sepals to estimate the orientation of tomatoes
Precise detection and localization are prerequisites for intelligent harvesting, while fruit size and weight estimation are key to intelligent orchard management. In commercial banana orchards, it is necessary to manage the growth and weight of banana bunches so that they can be harvested in time and prepared for transportation according to their different maturity levels. In this study, in order to reduce management costs and labor dependence, and obtain non-destructive weight estimation, we propose a method for localizing and estimating banana bunches using RGB-D images. First, the color image is detected through the YOLO-Banana neural network to obtain two-dimensional information about the banana bunches and stalks. Then, the three-dimensional coordinates of the central point of the banana stalk are calculated according to the depth information, and the banana bunch size is obtained based on the depth information of the central point. Finally, the effective pixel ratio of the banana bunch is presented, and the banana bunch weight estimation model is statistically analyzed. Thus, the weight estimation of the banana bunch is obtained through the bunch size and the effective pixel ratio. The R2 value between the estimated weight and the actual measured value is 0.8947, the RMSE is 1.4102 kg, and the average localization error of the central point of the banana stalk is 22.875 mm. The results show that the proposed method can provide bunch size and weight estimation for the intelligent management of banana orchards, along with localization information for banana-harvesting robots.
Why it matches plant phenotyping methodsRGB-D画像からバナナ房のサイズと重量という植物器官形質を推定する手法を開発・評価しており、収穫ロボット用の局在化にとどまらないため。
abstractwe propose a method for localizing and estimating banana bunches using RGB-D images
Greenhouse production of fruits and vegetables in developed countries is challenged by labour scarcity and high labour costs. Robots offer a good solution for sustainable and cost-effective production. Acquiring accurate spatial information about relevant plant parts is vital for successful robot operation. Robot perception in greenhouses is challenging due to variations in plant appearance, viewpoints, and illumination. This paper proposes a keypoint-detection-based method using data from an RGB-D camera to estimate the 3D pose of peduncle nodes, which provides essential information to harvest the tomato bunches. Specifically, this paper proposes a method that detects four anatomical landmarks in the colour image and then integrates 3D point-cloud information to determine the 3D pose. A comprehensive evaluation was conducted in a commercial greenhouse to gain insight into the performance of different parts of the method. The results showed: (1) high accuracy in object detection, achieving an Average Precision (AP) of [email protected]=0.96; (2) an average Percentage of Detected Joints (PDJ) of the keypoints of [email protected] = 94.31%; and (3) 3D pose estimation accuracy with mean absolute errors (MAE) of 11o and 10o for the relative upper and lower angles between the peduncle and main stem, respectively. Furthermore, the capability to handle variations in viewpoint was investigated, demonstrating the method was robust to view changes. However, canonical and higher views resulted in slightly higher performance compared to other views. Although tomato was selected as a use case, the proposed method has the potential to be applied to other greenhouse crops, such as pepper, after fine-tuning.
Why it matches plant phenotyping methodsトマトの器官ランドマークを画像・点群から抽出し、ペドンクル節の3D姿勢という植物形態形質を推定する手法を開発・評価しており、収穫ロボットへの応用を超えて表現型取得が中心である。
abstractThis paper proposes a keypoint-detection-based method using data from an RGB-D camera to estimate the 3D pose of peduncle nodes
Abstract Cucumis melo L., commonly known as melon, is a crucial horticultural crop. The selection and breeding of superior melon germplasm resources play a pivotal role in enhancing its marketability. However, current methods for melon appearance phenotypic analysis rely primarily on expert judgment and intricate manual measurements, which are not only inefficient but also costly. Therefore, to expedite the breeding process of melon, we analyzed the images of 117 melon varieties from two annual years utilizing artificial intelligence (AI) technology. By integrating the semantic segmentation model Dual Attention Network (DANet), the object detection model RTMDet, the keypoint detection model RTMPose, and the Mobile-Friendly Segment Anything Model (MobileSAM), a deep learning algorithm framework was constructed, capable of efficiently and accurately segmenting melon fruit and pedicel. On this basis, a series of feature extraction algorithms were designed, successfully obtaining 11 phenotypic traits of melon. Linear fitting verification results of selected traits demonstrated a high correlation between the algorithm-predicted values and manually measured true values, thereby validating the feasibility and accuracy of the algorithm. Moreover, cluster analysis using all traits revealed a high consistency between the classification results and genotypes. Finally, a user-friendly software was developed to achieve rapid and automatic acquisition of melon phenotypes, providing an efficient and robust tool for melon breeding, as well as facilitating in-depth research into the correlation between melon genotypes and phenotypes.
Why it matches plant phenotyping methodsメロン果実・果梗を画像から分割し、特徴抽出によって11形質を推定する深層学習フレームワークを開発・検証し、ソフトウェア化しているため、植物フェノタイピング手法が中心である。
abstractBy integrating the semantic segmentation model Dual Attention Network (DANet), the object detection model RTMDet, the keypoint detection model RTMPose, and the Mobile-Friendly Segment Anything Model (MobileSAM), a deep learning algorithm framework was constructed, capable of efficiently and accurately segmenting melon fruit and pedicel.
Reproduction assets found保存済みの本文根拠を更新済みルールで再検証し、公開資産1件を確認しました。Code · publicAdditionally, we have developed a simple melon phenotypic traits extraction software, which can be downloaded via https://github.com/hongbinz13/Melon-Phenotype-Extractor/releases/tag/software .Open asset ↗https://github.com/hongbinz13/Melon-Phenotype-Extractor · Melon-Phenotype-Extractorlines:109-139Code / dataset availability confirmedEurope PMC · checked 7 Sept 2026
Image segmentation is commonly used to estimate the location and shape of plants and their external structures. Segmentation masks are then used to localize landmarks of interest and compute other geometric features that correspond to the plant's phenotype. Despite its prevalence, segmentation-based approaches are laborious (requiring extensive annotation to train) and error-prone (derived geometric features are sensitive to instance mask integrity). Here, we present a segmentation-free approach that leverages deep learning-based landmark detection and grouping, also known as pose estimation. We use a tool originally developed for animal motion capture called SLEAP (Social LEAP Estimates Animal Poses) to automate the detection of distinct morphological landmarks on plant roots. Using a gel cylinder imaging system across multiple species, we show that our approach can reliably and efficiently recover root system topology at high accuracy, few annotated samples, and faster speed than segmentation-based approaches. In order to make use of this landmark-based representation for root phenotyping, we developed a Python library ( sleap-roots ) for trait extraction directly comparable to existing segmentation-based analysis software. We show that pose-derived root traits are highly accurate and can be used for common downstream tasks including genotype classification and unsupervised trait mapping. Altogether, this work establishes the validity and advantages of pose estimation-based plant phenotyping. To facilitate adoption of this easy-to-use tool and to encourage further development, we make sleap-roots , all training data, models, and trait extraction code available at: https://github.com/talmolab/sleap-roots and https://osf.io/k7j9g/.
Why it matches plant phenotyping methods根系のランドマーク検出・形状復元・形質抽出を行う深層学習ベースの植物フェノタイピング手法を開発・検証し、専用ライブラリも提供しているため。
abstractHere, we present a segmentation-free approach that leverages deep learning-based landmark detection and grouping, also known as pose estimation.
Reproduction assets foundThe authors explicitly make all paper-specific assets public: the sleap-roots trait-extraction codebase on GitHub, a separate repository with figure-replication code, and an OSF deposit containing labeled training data, trained pose-estimation models, and analysis files for the root phenotyping measurements.Code · publicthe specific code utilized for replicating the figures presented in this study can be found in a separate GitHub repository here: https://github.com/talmolab/Berrigan_et_al_sleap-rootsOpen asset ↗talmolab/Berrigan_et_al_sleap-roots · Berrigan_et_al_sleap-rootslines:485-526Dataset · publicThe datasets generated and/or analyzed during the current study are available in the Open Science Framework (OSF) repository. This includes the labeled data, predictive models, and analysis files, which can be accessed via the following link: https://osf.io/k7j9g/Open asset ↗osf.io/k7j9g · k7j9glines:485-526Plant phenotyping relevance match · UnverifiedEurope PMC · checked 15 Sept 2026
Published1 Apr 2024Computers and Electronics in Agriculture.
Apple is a highly valued specialty crop in the U.S. Green fruit thinning is an important operation of apple production, which is the removal of excess fruitlets in the early summer. The task ensures that remaining fruits at harvest time grow to have good size and quality while reducing the risk of biennial bearing. Current methods of thinning include hand, chemical, and mechanical. However, hand thinning generally requires a large labor force to implement, chemical thinning is non-selective and dependent on timing and weather during application, and mechanical thinning is also non-selective and destructive. A robotic green fruit thinning system could possibly be implemented that does not exhibit the drawbacks of current methods. A vision system is an essential component for a robotic green fruit thinning system that is responsible for green fruit detection and segmentation, decision-making on which fruit to remove, and environment reconstruction for path planning. This study took the first step towards developing a vision system for robotic green fruit thinning. First, green fruit and stem instance segmentation was applied using Mask R-CNN. Then, green fruit and stem orientation estimation was applied using Principal Component Analysis (PCA). Average precision scores for green fruit and stem segmentation on all mask sizes were 83.4% and 38.9%, respectively, whereas these increased to 91.3% and 67.7% if only considering the fruits and stems with mask sizes greater than 32² pixels. Green fruit orientation estimation with correction made 89.3% and 75.5% of estimates accurate within 30° of actual orientations for ground-truth and segmentation-generated masks, respectively. Performances respectively were 97.4% and 84.0% when only unoccluded masks are considered. Orientation correction resulted in considerable improvements in all cases of green fruit orientation estimation, with the greatest improvement seen on unoccluded ground truth masks where estimates accurate within 30° of ground truth orientations increased by 23.9%. Stem orientation estimation achieved very high accuracies with corresponding scores of 99.8% and 99.7%. The outcomes provided guideline information for developing a robust machine vision system for robotic green fruit thinning.
Why it matches plant phenotyping methodsリンゴ果実・茎のセグメンテーションと方向推定という、植物器官の形態形質を抽出する画像解析手法を開発・評価しており、ロボット作業の位置決めだけでなく再利用可能な表現型推定が中心である。
abstractThis study took the first step towards developing a vision system for robotic green fruit thinning.
Agricultural production is facing severe challenges in the next decades induced by climate change and the need for sustainability, reducing its impact on the environment. Advancements in field management through non-chemical weeding by robots in combination with monitoring of crops by autonomous unmanned aerial vehicles (UAVs) and breeding of novel and more resilient crop varieties are helpful to address these challenges. The analysis of plant traits, called phenotyping, is an essential activity in plant breeding, it however involves a great amount of manual labor. With this paper, we address the problem of automatic fine-grained organ-level geometric analysis needed for precision phenotyping. As the availability of real-world data in this domain is relatively scarce, we propose a novel dataset that was acquired using UAVs capturing high-resolution images of a real breeding trial containing 48 plant varieties and therefore covering great morphological and appearance diversity. This enables the development of approaches for autonomous phenotyping that generalize well to different varieties. Based on overlapping high-resolution images from multiple viewing angles, we compute photogrammetric dense point clouds and provide detailed and accurate point-wise labels for plants, leaves, and salient points as the tip and the base. Additionally, we include measurements of phenotypic traits performed by experts from the German Federal Plant Variety Office on the real plants, allowing the evaluation of new approaches not only on segmentation and keypoint detection but also directly on the downstream tasks. The provided labeled point clouds enable fine-grained plant analysis and support further progress in the development of automatic phenotyping approaches, but also enable further research in surface reconstruction, point cloud completion, and semantic interpretation of point clouds.
Why it matches plant phenotyping methods植物の器官レベル表現型解析を目的とするUAV画像由来の点群データセットで、植物・葉のラベルと専門家による形質測定を提供し、自動フェノタイピング手法の評価を可能にするため、方法論が中心である。
abstractwe propose a novel dataset that was acquired using UAVs capturing high-resolution images of a real breeding trial
Rapid and precise tree Diameter at Breast Height (DBH) measurement is pivotal in forest inventories. While the recent advancements in LiDAR and Structure from Motion (SFM) technologies have paved the way for automated DBH measurements, the significant equipment costs and the complexity of operational procedures continue to constrain the ubiquitous adoption of these technologies for real-time DBH assessments. In this research, we introduce KAN-Forest, a real-time DBH measurement and key point localization algorithm utilizing RGB-D (Red, Green, Blue-Depth) imaging technology. Firstly, we improved the YOLOv5-seg segmentation module with a Channel and Spatial Attention (CBAM) module, augmenting its efficiency in extracting the tree’s edge features in intricate forest scenarios. Subsequently, we devised an image processing algorithm for real-time key point localization and DBH measurement, leveraging historical data to fine-tune current frame assessments. This system facilitates real-time image data upload via wireless LAN for immediate host computer processing. We validated our approach on seven sample plots, achieving bbAP50 and segAP50 scores of: 90.0%(+3.0%), 90.9%(+0.9%), respectively with the improved YOLOv5-seg model. The method exhibited a DBH estimation RMSE of 17.61∼54.96 mm (R2=0.937), and secured 78% valid DBH samples at a 59 FPS. Our system stands as a cost-effective, portable, and user-friendly alternative to conventional forest survey techniques, maintaining accuracy in real-time measurements compared to SFM- and LiDAR-based algorithms. The integration of WLAN and its inherent scalability facilitates deployment on Unmanned Ground Vehicles (UGVs) to improve the efficiency of forest inventory. We have shared the algorithms and datasets on Github for peer evaluations.
Why it matches plant phenotyping methodsRGB-D画像とアルゴリズムを用いて樹木のDBHという明示的な形態形質をリアルタイム推定し、精度検証とシステム実装を行った研究であり、植物フェノタイピング手法が中心である。
abstractwe introduce KAN-Forest, a real-time DBH measurement and key point localization algorithm utilizing RGB-D (Red, Green, Blue-Depth) imaging technology.
Reproduction assets foundThe authors explicitly state that the code used in this DBH measurement research is publicly available on GitHub (KAN-Forest repository), making it a paper-specific, public, actionable code asset. The phenotype/trait datasets (DBH measurements and forest images) are only available upon request from the corresponding作者,Code · publicwe have made
the code used in this research available on GitHub at: https://github.com/CharmingZh/KAN-ForestOpen asset ↗CharmingZh/KAN-Forestpdf-page:27 lines:1-59Code / dataset availability confirmedOpenAlex · Europe PMC · bioRxiv · checked 7 Sept 2026
Abstract Image segmentation is commonly used to estimate the location and shape of plants and their external structures. Segmentation masks are then used to localize landmarks of interest and compute other geometric features that correspond to the plant’s phenotype. Despite its prevalence, segmentation-based approaches are laborious (requiring extensive annotation to train), and error-prone (derived geometric features are sensitive to instance mask integrity). Here we present a segmentation-free approach which leverages deep learning-based landmark detection and grouping, also known as pose estimation. We use a tool originally developed for animal motion capture called SLEAP (Social LEAP Estimates Animal Poses) to automate the detection of distinct morphological landmarks on plant roots. Using a gel cylinder imaging system across multiple species, we show that our approach can reliably and efficiently recover root system topology at high accuracy, few annotated samples, and faster speed than segmentation-based approaches. In order to make use of this landmark-based representation for root phenotyping, we developed a Python library ( sleap-roots ) for trait extraction directly comparable to existing segmentation-based analysis software. We show that landmark-derived root traits are highly accurate and can be used for common downstream tasks including genotype classification and unsupervised trait mapping. Altogether, this work establishes the validity and advantages of pose estimation-based plant phenotyping. To facilitate adoption of this easy-to-use tool and to encourage further development, we make sleap-roots , all training data, models, and trait extraction code available at: https://github.com/talmolab/sleap-roots and https://osf.io/k7j9g/ .
Why it matches plant phenotyping methods植物根の形態ランドマークをポーズ推定で検出し、根系形質を抽出する手法とソフトウェアを開発・検証した研究であり、植物フェノタイピング手法が中心である。
abstractHere we present a segmentation-free approach which leverages deep learning-based landmark detection and grouping, also known as pose estimation.
Reproduction assets foundThe paper makes its root phenotyping assets public: labeled training data, trained SLEAP models, and analysis files on OSF, the sleap-roots trait-extraction codebase on GitHub, and a separate figure-replication code repository.Dataset · publicThe datasets generated and/or analyzed during the current study are available in the Open Science Framework (OSF) repository. This includes the labeled data, predictive models, and analysis files which can be accessed via the following link: https://osf.io/k7j9g/ .Open asset ↗osf.io/k7j9glines:752-811Code · publicAdditionally, the specific code utilized for replicating the figures presented in this study can be found in a separate GitHub repository here: https://github.com/talmolab/Berrigan_et_al_sleap-roots .Open asset ↗talmolab/Berrigan_et_al_sleap-rootslines:752-811Plant phenotyping relevance match · UnverifiedEurope PMC · checked 15 Sept 2026
Published1 Nov 2023Computers and Electronics in Agriculture.
It is necessary to recognize the tomato pollination features for the designing demand of intelligent and precise tomato supplementary pollination equipment. Mentioned pollination features include flower opening state and the three-dimensional position and pose of flower anther. Tomato flower pollination features recognition method is designed in this paper based on the deep learning full opened flower recognition model and binocular template matching three-dimensional information recognition method. First of all, a binocular stereo vision system is built to acquire the tomato flower images in greenhouse with natural lighting. Which can help vision system avoid the impact of inconsistent light intensity on image recognition. The acquired images were equalized with three groups parameters to labeled images. And the improved MC-AlexNet deep learning model is established to recognize the full opened tomato flowers in the left image of image pair acquired with binocular vision system. Then, the template is created with the full opened flower recognition result of deep learning model in left image based on gray value and deformation template matching method. And the template matching is established to recognize the corresponding full opened flowers in the right image of image pair. Finally, with the template matching result, the anther segmentation is conducted to calculate three-dimensional position and pose of anther. The experimental results show that the accuracy of full opened tomato flowers recognition model is 96.23%. The average position recognition deviation of anther is 5.94 mm. And the average anther pose recognition deviation in three plane is 6.24° for the images that anthers can be observed. And the average time consuming is about 143.18 ms per image pair. It can be concluded that the method of binocular template matching established in this paper can fulfill the demand of supplementary pollination equipment design, and the research result lays a foundation for designing and improvement of tomato supplementary pollination equipment in greenhouse.
Why it matches plant phenotyping methodsトマト花の開花状態と葯の三次元位置・姿勢という植物器官形質を、深層学習・両眼ステレオ・テンプレートマッチングで取得する方法を開発・評価しており、フェノタイピング手法が中心である。
abstractTomato flower pollination features recognition method is designed in this paper based on the deep learning full opened flower recognition model and binocular template matching three-dimensional information recognition method.
RootClassificationMorphology / geometry measurementPose / keypoint estimationRoot system architecture
Image segmentation is commonly used to estimate the location and shape of plants and their external structures. Segmentation masks are then used to localize landmarks of interest and compute other geometric features that correspond to the plant’s phenotype. Despite its prevalence, segmentation-based approaches are laborious (requiring extensive annotation to train), and error-prone (derived geometric features are sensitive to instance mask integrity). Here we present a segmentation-free approach which leverages deep learning-based landmark detection and grouping, also known as pose estimation. We use a tool originally developed for animal motion capture called SLEAP (Social LEAP Estimates Animal Poses) to automate the detection of distinct morphological landmarks on plant roots. Using high-throughput phenotyping method Root Architecture 3-D Imaging Cylinder (RADICYL) across multiple species, we show that our approach can reliably and efficiently recover root system topology at greater accuracy, faster speed, and with fewer annotated samples than segmentation-based approaches. In order to make use of this landmark-based representation for root phenotyping, we developed a Python library ( sleap-roots ) for trait extraction directly comparable to existing segmentation-based analysis software. We show that landmark-derived root traits are highly accurate and can be used for common downstream tasks including genotype classification and unsupervised trait mapping. Altogether, this work establishes the validity and advantages of pose estimation-based plant phenotyping. To facilitate adoption of this easy-to-use tool and to encourage further development, we make sleap-roots , all training data, models, and trait extraction code available at: https://github.com/talmolab/sleap-roots .
Why it matches plant phenotyping methods植物根のランドマーク検出による表現型抽出法を開発・検証し、ソフトウェアと学習データも提供しているため、方法が中心的です。
abstractHere we present a segmentation-free approach which leverages deep learning-based landmark detection and grouping, also known as pose estimation.
Reproduction assets foundThe authors explicitly state they make their sleap-roots Python library, all training data, trained models, and trait extraction code publicly available on GitHub, directly supporting this paper's root pose-estimation phenotyping analysis.Code · publice classification and unsupervised
trait mapping. Altogether, this work establishes the validity and advantages of pose estimation-based plant phenotyping. To
facilitate adoption of this easy-to-use tool and to encourage further development, we make sleap-roots , all training data,
models, and trait extraction code available at: https://github.com/talmolab/sleap-roots.
Elizabeth M. Berrigan, Lin Wang, Hannah Carrillo, Kimberly Echegoyen, Mikayla Kappes, Jorge Torres,
Angel Ai-Perreira, Erica McCoy, Emily Shane, Charles Copeland, Lauren Ragel, Charidimos Georgousakis,
Sanghwa Lee, Dawn Reynolds, Avery Talgo, Juan Gonzalez, Ling Zhang, Ashish Rajurkar, Michel Ruiz,
Erin Daniels, Liezl Maree, SOpen asset ↗talmolab/sleap-rootspdf-layout-page:1 lines:1-47Plant phenotyping relevance match · UnverifiedEurope PMC · checked 15 Sept 2026
Published1 Aug 2023Computers and Electronics in Agriculture.
Automated fruit harvesting is promising research in the development of agricultural modernization. However, the complex and non-structural orchard environment is extremely challenging. In order to meet the needs of different end-effectors and to improve the success rate of automatic fruit harvesting, it is critical to perform fruit pose estimation before picking operations. In this study, a citrus pose estimation method through a single RGB image is introduced. The rotation of the citrus pose is defined as a vector that passes through the center of the fruit, which is perpendicular to the plane where the fruit navel point is located. Simply speaking, a multi-task learning model named FPENet is proposed to simultaneously locate the fruit navel point and predict the fruit rotation vector. And a hyperparameter is introduced in the loss function to achieve the simultaneous convergence of multiple tasks. In addition, this paper designs a 2D image annotation tool and constructs a citrus pose dataset, which contributes to model training and also the algorithm evaluation. In the experiment, we evaluate and analyze each module of the proposed network structure, and verify its performance on a harvesting robot. The experimental results show that the FPENet achieves an 88.92 AP score on fruit navel point detection, and 11.13° on the average error of the rotation vector. Over 90% of rotation vectors have an angular error of less than 22.5°. The harvesting success rate is 79.79%. This study offers a new idea for fruit pose estimation and provides the possibility and foundation for estimating fruit pose with a 2D image input.
Why it matches plant phenotyping methodsRGB画像から果実のへそ位置と回転ベクトル（果実姿勢）を推定する手法を開発し、データセットと注釈ツールも構築している。収穫対象の単なる検出・位置特定を超えて、再利用可能な果実器官の姿勢形質を抽出するため、中心的なフェノタイピング手法と判断する。
abstractIn this study, a citrus pose estimation method through a single RGB image is introduced.
The precise detection and positioning of tea buds are among the major issues in tea picking automation. In this study, a novel algorithm for detecting tea buds and estimating their poses in a field environment was proposed by using a depth camera. This algorithm introduces some improvements to the YOLOv5l architecture. A Coordinate Attention Mechanism (CAM) was inserted into the neck part to accurately position the elements of interest, a BiFPN was used to enhance the small object detection ability, and a GhostConv module replaced the original Conv module in the backbone to reduce the model size and speed up model inference. After testing, the proposed detection model achieved an mAP of 85.2%, a speed of 87.71 FPS, a parameter number of 29.25 M, and a FLOPs value of 59.8 G, which are all better than those achieved with the original model. Next, an optimal pose-vertices search method (OPVSM) was developed to estimate the pose of tea by constructing a graph model to fit the pointcloud. This method could accurately estimate the poses of tea buds, with an overall accuracy of 90%, and it was more flexible and adaptive to the variations in tea buds in terms of size, color, and shape features. Additionally, the experiments demonstrated that the OPVSM could correctly establish the pose of tea buds through pointcloud downsampling by using voxel filtering with a 2 mm × 2 mm × 1 mm grid, and this process could effectively reduce the size of the pointcloud to smaller than 800 to ensure that the algorithm could be run within 0.2 s. The results demonstrate the effectiveness of the proposed algorithm for tea bud detection and pose estimation in a field setting. Furthermore, the proposed algorithm has the potential to be used in tea picking robots and also can be extended to other crops and objects, making it a valuable tool for precision agriculture and robotic applications.
Why it matches plant phenotyping methods茶芽の検出と3D姿勢推定という植物器官の形態・状態を取得する画像／深度センシング手法を開発し、精度・速度を評価しており、フェノタイピング手法が中心である。
abstracta novel algorithm for detecting tea buds and estimating their poses in a field environment was proposed by using a depth camera.
Accurate detection of tea shoots and precise location of picking points are prerequisites for automated, intelligent and accurate tea picking. A method was developed for the detection of tea shoots and key points and the localisation of picking points in complex environments. Images of four types of tea shoots were collected from multiple fields of view in a tea plantation over two months and labelling criteria were established. The YOLO-Tea model was developed based on the YOLOv5 network model, which uses a content-based upsampling operator (CARAFE) with a larger field of perception to implement the tea shoot feature upsampling operation, adds a convolutional attention mechanism module (CBAM) to focus the model on both channel and spatial dimensions to detect and localise important areas of tea shoots in a large field of view. The Bottleneck Transformers module was used to inject global self-focus for residuals to create long-distance dependencies on the tea shot feature images, and a six-point landmark regression head was added. The experimental results demonstrated that the YOLO-Tea model improved the mean Average Precision (mAP) value of tea shoots and their key points by 5.26% compared to YOLOv5. Finally, we use image processing methods to locate picking point positions based on key point information during the model inference phase. This study has theoretical and practical implications for the detection of tea shoots and their key points, tea shoot alignment, phenotype identification, pose estimation and picking locations of premium teas in complex environments.
Why it matches plant phenotyping methods茶芽の検出・キーポイント推定・摘採点位置推定を目的とする画像ベース手法を開発し、YOLOv5との性能比較で検証している。ロボット摘採向けだが、単なる対象位置特定に留まらず、茶芽の形態・姿勢に関わる表現型推定を含むため中心的なフェノタイピング手法研究である。
abstractA method was developed for the detection of tea shoots and key points and the localisation of picking points in complex environments.
Conformation traits impact gait and performance, and as a result vary based on intended use of the horse. For example, a horse with longer legs can cover a greater distance per stride than one with shorter legs. Limb length, therefore, is advantageous for horses expected to travel long distances for extended periods of time but may not be for those who go short distances at top speeds. Fragile Foal Syndrome (FFS) in horses results from a sequence change in PLOD1, the same gene that causes Ehlers Danlos Syndrome type IV (EDS-IV) in humans, a condition that results in longer limbs. Based on the phenotype observed in EDS, we hypothesize that FFS in the carrier state may impact extremity length including the neck, forelimbs, and hindlimbs in horses. To measure conformation parameters, we first digitally labeled specific anatomical points on a video recording of a trotting horse using the software package DeepLabCut (DLC). Representative frames for each horse at the same point within the stride were extracted from the video. The X, Y coordinates of each anatomical label were used to quantify conformation geometry. Frame extraction was done with a custom designed graphical interface that ensured the user could identify the ideal frame matching an identical body position across all horses examined. X,Y coordinate output files were thenanalyzed with a custom gait parameter pipeline written in MatLab to calculate the neck, forelimb, and hindlimb length. Each measurement was scaled to account for the varying distance horses were from the camera. Pulled tail hair was utilized to extract DNA and genotype each horse using PCR-RFLP, then followed by Sanger Sequencing for confirmation. Conformational measures were compared between genotype categories (wild type vs, carrier) in JMP Pro. Preliminary results showed that FFS carriers (n = 7) had significantly longer forelimb (P = 0.0344, t-test), hindlimb (P = 0.0258, t-test), and neck (P = 0.0013, t-test) lengths compared with wild type horses (n = 7). Future work will increase the overall sample size and number of frames captured per individual to increase statistical power and allow investigation of more conformation traits including static joint angles and back lengths. Improved understanding of conformational phenotypes that may result from the FFS carrier genotype will inform future studies of equine physiology and connective tissue disorders, leading to better decision making for breeding, sales and training, and preventing injury and loss of foals to this disease.
Why it matches plant phenotyping methods動画画像とDeepLabCut、カスタムGUI・解析パイプラインを用いて馬体の肢長・頸長などの形態形質を定量化しており、表現型取得・抽出法が研究の中心的要素である。
abstractTo measure conformation parameters, we first digitally labeled specific anatomical points on a video recording of a trotting horse using the software package DeepLabCut (DLC).
The dormant pruning of jujube is a labor-intensive and time-consuming activity in the production and management of jujube orchards, which mainly depends on manual operation. Automatic pruning using robots could be a better way to solve the shortage of skilled labor and improve efficiency. In order to realize automatic pruning of jujube trees, a method of pruning point identification based on skeleton information is presented. This study used an RGB-D camera to collect multi-view information on jujube trees and built a complete point cloud information model of jujube trees. The space colonization algorithm acts on the global point cloud to generate the skeleton of jujube trees. The iterative relationship between skeleton points was represented by constructing a directed graph. The proposed skeleton analysis algorithm marked the skeleton as the trunk, the primary branches, and the lateral branches and identified the pruning points under the guidance of pruning rules. Finally, the visual model of the pruned jujube tree was established through the skeleton information. The results showed that the registration errors of individual jujube trees were less than 0.91 cm, and the average registration error was 0.66 cm, which provided a favorable database for skeleton extraction. The skeleton structure extracted by the space colonization algorithm had a high degree of coincidence with jujube trees, and the identified pruning points were all located on the primary branches of jujube trees. The study provides a method to identify the pruning points of jujube trees and successfully verifies the validity of the pruning points, which can provide a reference for the location of the pruning points and visual research basis for automatic pruning.
Why it matches plant phenotyping methodsRGB-D点群からナツメ樹の樹幹・一次枝・側枝の骨格を抽出し、剪定点を推定する画像・計算手法が研究の中心であり、植物形態・樹体構造の表現型取得として妥当。
abstractThis study used an RGB-D camera to collect multi-view information on jujube trees and built a complete point cloud information model of jujube trees.
Plant phenotyping relevance match · UnverifiedOpenAlex · Europe PMC · checked 14 Sept 2026
A high-throughput image analysis pipeline was developed to facilitate root phenotyping by reducing time-consuming labeling while maintaining phenotyping accuracy. This pipeline leverages a deep learning-based tool named SLEAP (SLEAP Estimates Animal Poses) which is designed to automate the detection of distinct morphological landmarks. By training SLEAP to detect the root branch points, tips, and midline of each root imaged in a gel cylinder, we were able to robustly and efficiently recover the root system geometry. We trained models to identify these landmarks on primary, lateral, and seminal roots across a range of crop plants, including soybean, rice, canola, and pennycress. We find that our SLEAP models are robust across genotypes and experiments, enabling automated root system quantification at the rate of hundreds of plants per hour. Using predictions of root landmark locations, we developed Python-based pipelines to extract phenotypic traits, including tip depths, root lengths, convex hulls, root angles, measures of curviness, and lateral root distribution (available at https://github.com/talmolab/sleap-roots). In order to extract meaningful patterns from this high-dimensional description of plant phenotypes, we use machine learning-based methods for dimensionality reduction and manifold embedding, allowing us to capture the statistical structure of root phenotypes present in our screens. In future work, we will use these quantitative phenotypic traits as a predictor for root system traits that enhance carbon sequestration capabilities in genome-wide association studies.
Why it matches plant phenotyping methods深層学習による根のランドマーク検出と画像解析パイプラインを開発し、根系形態形質を自動抽出する研究であり、植物フェノタイピング手法が中心である。
abstractA high-throughput image analysis pipeline was developed to facilitate root phenotyping by reducing time-consuming labeling while maintaining phenotyping accuracy.
The maize detasseling process was gradually automated with the continuous promotion of the maize detasseling machine. Nevertheless, some problems are gradually becoming more prominent for it. Most maize detasseling machine has a simple maize tassels detection system. That is, the height of the crop was identified, and not the specific information on the maize tassels, leading to low detasseling precision and a high leaf injury rate during maize detasseling machine working. Aiming at these issues, this paper proposed a novel method for the maize tassel pose estimation based on computer vision and oriented object detection. Specifically, a two-step framework for maize tassel pose estimation is developed. Firstly, the maize plant is captured from above, then the maize tassels and the second leaf’s vein are posed in the horizontal plane, and their poses are matched to the oriented bounding box. Second, an oriented bounding box is generated according to the Oriented R-CNN model detection. Since maize tassels in the field differ in size and the morphological color of different growth stages, estimating the maize tassel pose accurately by the angle of the oriented bounding box alone is unfeasible. Therefore, these pixels of the oriented bounding box were put into the Look Twice module to extract the critical information about maize tassels, which can accurately determine the final pose of the maize tassels. Finally, evaluation metrics on the test set indicate the proposed method performed with correct maize tassels pose estimation rate of 88.56% and 29.57 Giga Floating-point Operations (GFLOPs), which indicated that the feasibility of maize tassel poses estimation using the proposed method. This study provides the possibility and foundation for precise maize detasseling in maize detasseling machines.
Why it matches plant phenotyping methodsトウモロコシ雄穂の姿勢という植物器官形質を画像から推定する手法を開発・評価しており、単なる位置検出を超えた形質抽出が中心である。
abstractthis paper proposed a novel method for the maize tassel pose estimation based on computer vision and oriented object detection.
To guide grape picking robots to recognize and classify the grapes with different maturity quickly and accurately in the complex environment of the orchard, and to obtain the spatial position information of the grape clusters, an algorithm of grape maturity detection and visual pre-positioning based on improved YOLOv4 is proposed in this study. The detection algorithm uses Mobilenetv3 as the backbone feature extraction network, uses deep separable convolution instead of ordinary convolution, and uses the h-swish function instead of the swish function to reduce the number of model parameters and improve the detection speed of the model. At the same time, the SENet attention mechanism is added to the model to improve the detection accuracy, and finally the SM-YOLOv4 algorithm based on improved YOLOv4 is constructed. The experimental results of maturity detection showed that the overall average accuracy of the trained SM-YOLOv4 target detection algorithm under the verification set reached 93.52%, and the average detection time was 10.82 ms. Obtaining the spatial position of grape clusters is a grape cluster pre-positioning method based on binocular stereo vision. In the pre-positioning experiment, the maximum error was 32 mm, the mean error was 27 mm, and the mean error ratio was 3.89%. Compared with YOLOv5, YOLOv4-Tiny, Faster_R-CNN, and other target detection algorithms, which have greater advantages in accuracy and speed, have good robustness and real-time performance in the actual orchard complex environment, and can simultaneously meet the requirements of grape fruit maturity recognition accuracy and detection speed, as well as the visual pre-positioning requirements of grape picking robots in the orchard complex environment. It can reliably indicate the growth stage of grapes, so as to complete the picking of grapes at the best time, and it can guide the robot to move to the picking position, which is a prerequisite for the precise picking of grapes in the complex environment of the orchard.
Why it matches plant phenotyping methodsブドウ果実の成熟度という植物状態を画像から推定するYOLO法を開発・検証しており、成熟度検出の精度・速度評価が研究の中心である。位置推定も含むが、成熟度フェノタイピング手法が明確に含まれる。
abstractan algorithm of grape maturity detection and visual pre-positioning based on improved YOLOv4 is proposed in this study.
Intelligent detection and localization of mature citrus fruits is a critical challenge in developing an automatic harvesting robot. Variable illumination conditions and different occlusion states are some of the essential issues that must be addressed for the accurate detection and localization of citrus in the orchard environment. In this paper, a novel method for the detection and localization of mature citrus using improved You Only Look Once (YOLO) v5s with binocular vision is proposed. First, a new loss function (polarity binary cross-entropy with logit loss) for YOLO v5s is designed to calculate the loss value of class probability and objectness score, so that a large penalty for false and missing detection is applied during the training process. Second, to recover the missing depth information caused by randomly overlapping background participants, Cr-Cb chromatic mapping, the Otsu thresholding algorithm, and morphological processing are successively used to extract the complete shape of the citrus, and the kriging method is applied to obtain the best linear unbiased estimator for the missing depth value. Finally, the citrus spatial position and posture information are obtained according to the camera imaging model and the geometric features of the citrus. The experimental results show that the recall rates of citrus detection under non-uniform illumination conditions, weak illumination, and well illumination are 99.55%, 98.47%, and 98.48%, respectively, approximately 2-9% higher than those of the original YOLO v5s network. The average error of the distance between the citrus fruit and the camera is 3.98 mm, and the average errors of the citrus diameters in the 3D direction are less than 2.75 mm. The average detection time per frame is 78.96 ms. The results indicate that our method can detect and localize citrus fruits in the complex environment of orchards with high accuracy and speed. Our dataset and codes are available at https://github.com/AshesBen/citrus-detection-localization.
Why it matches plant phenotyping methods収穫ロボット向けの位置検出が主目的だが、果実形状・姿勢・3D直径を画像から抽出し、精度を検証する技術開発が中心であり、再利用可能な植物器官形質の推定に該当する。
abstracta novel method for the detection and localization of mature citrus using improved You Only Look Once (YOLO) v5s with binocular vision is proposed.
Reproduction assets foundThe authors explicitly state that their citrus image dataset (4855 binocular image groups with depth maps) and analysis code are publicly available on GitHub, matching the allowed URL.Dataset · publicnt occlusion conditions in natural orchards. Future work will focus on few-shot learning and reduce the number of citrus fruits in the training dataset to improve citrus detection and localization.
Data availability statement
The original contributions presented in this study are publicly available. This data can be found here: https://github.com/AshesBen/citrus-detection-localization .
Author contributions
All authors contributed to the method and result of the study, dataset generation, model training and testing, analysis of results, and the drafting, revising, and approving of the contents of the manuscript.
Funding
We acknowledged support from the Natural Science Foundation of GuangdongOpen asset ↗AshesBen/citrus-detection-localizationlines:335-356Code · public98 mm, and the average errors of the citrus diameters in the 3D direction are less than 2.75 mm. The average detection time per frame is 78.96 ms. The results indicate that our method can detect and localize citrus fruits in the complex environment of orchards with high accuracy and speed. Our dataset and codes are available at https://github.com/AshesBen/citrus-detection-localization .
Keywords: citrus detection, citrus localization, binocular vision, YOLO v5s, loss function
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 Jun 18; Accepted 2022 Jul 12; Collection date 2022.
IntroductionOpen asset ↗AshesBen/citrus-detection-localizationlines:1-28Plant phenotyping relevance match · UnverifiedCrossref · checked 15 Sept 2026
Monitoring of fields and breeding plots is critical for farmers, plant scientists, and breeders. In this process, a key objective is to assess and monitor the growth stages together with the number of individual plants on the field. Traditionally, this in-field assessment is performed manually and thus is limited in temporal and spatial throughput. In contrast, vision-based systems offer the potential to assess these traits frequently in an automated fashion on a large scale. The primary target of these systems is to detect and segment each plant and its leaves since this information directly correlates to the growth stage and allows for detailed monitoring. In this letter, we address the problem of automated, instance-level plant monitoring in agricultural fields and breeding plots. We propose a vision-based approach to perform a joint instance segmentation of crop plants and leaves in breeding plots. We develop a convolutional neural network to determine the position of specific plant keypoints and group pixels to detect individual leaf and plant instances. Finally, we provide a pixel-wise instance segmentation of each crop and its associated leaves based on orthorectified RGB images captured by UAVs. The experimental evaluation shows that our method outperforms state-of-the-art instance segmentation approaches such as Mask-RCNN on this task.
Why it matches plant phenotyping methodsUAV画像から個体・葉を自動検出・分割し、植物の成長段階や個体数評価に用いる手法を開発・評価しており、表現型取得が中心である。
abstractvision-based systems offer the potential to assess these traits frequently in an automated fashion on a large scale.
Non-destructive picking of fresh tomatoes is a delicate agronomical operation, based on comprehensive information about the plant organ, such as the location of stem, peduncle, and fruits. The matching between visual information supply and information demand from the agronomical technic is the key power to promote the picking robot from the laboratory to the field. The three-dimensional pose information, containing the location of each organ of the plant, can meet the demand of agronomical technic. It is the premise of precisely handling the cluster of fruits. In order to realize the fine tomato bunch harvesting operation in a bunch, this paper proposed a three-dimensional pose detection method for tomato bunch. The method, named Tomato Pose Method (TPM), is composed of a priori geometric model, a cascaded multi-task network, and a three-dimensional reconstruction process. Based on prior knowledge and agronomic technology, this prior geometric model comprehensively and flexibly describes the spatial location information of tomato bunch. The cascaded multi-task network is designed based on hourglass structure and transfer learning, which is suitable for bounding box and key point prediction of tomato bunches in complex environments. Finally, combining the prior geometric model and the spatial position information of each key point, the tomato bunch is reconstructed. Only a medium training dataset, containing 1800 RGBD images covering changing lighting, occlusion, and various poses, is needed for training. Its success rate of TPM on two-dimensional keypoint detection is 94.02%, the accuracy of 85.77% predicted points are at medium level. And 70.05% tomato bunch with multi-pose can be constructed. More importantly, this method only needs one RGBD image taken by a commercial camera to realize the three-dimensional reconstruction of a single-bunch scenario in 1.0 s, and a multi-bunch scenario in 2.0 s. It provides comprehensive information, and provides data basis for target positioning and path planning of picking robot, which makes the non-destructive harvesting possible.
Why it matches plant phenotyping methodsトマト果房の茎・花柄・果実の三次元位置と姿勢をRGB-D画像から抽出・再構成する手法を開発し、検出精度と再構成性能を評価している。収穫ロボット応用だが、再利用可能な植物器官形態の計測が中心である。
abstractThe three-dimensional pose information, containing the location of each organ of the plant, can meet the demand of agronomical technic.
Abstract Background In recent years, there has been an increase of interest in plant behaviour as represented by growth-driven responses. These are generally classified into nastic (internally driven) and tropic (environmentally driven) movements. Nastic movements include circumnutations, a circular movement of plant organs commonly associated with search and exploration, while tropisms refer to the directed growth of plant organs toward or away from environmental stimuli, such as light and gravity. Tracking these movements is therefore fundamental for the study of plant behaviour. Convolutional neural networks, as used for human and animal pose estimation, offer an interesting avenue for plant tracking. Here we adopted the Social LEAP Estimates Animal Poses (SLEAP) framework for plant tracking. We evaluated it on time-lapse videos of cases spanning a variety of parameters, such as: (i) organ types and imaging angles (e.g., top-view crown leaves vs. side-view shoots and roots), (ii) lighting conditions (full spectrum vs. IR), (iii) plant morphologies and scales (100 μm-scale Arabidopsis seedlings vs. cm-scale sunflowers and beans), and (iv) movement types (circumnutations, tropisms and twining). Results Overall, we found SLEAP to be accurate in tracking side views of shoots and roots, requiring only a low number of user-labelled frames for training. Top views of plant crowns made up of multiple leaves were found to be more challenging, due to the changing 2D morphology of leaves, and the occlusions of overlapping leaves. This required a larger number of labelled frames, and the choice of labelling “skeleton” had great impact on prediction accuracy, i.e., a more complex skeleton with fewer individuals (tracking individual plants) provided better results than a simpler skeleton with more individuals (tracking individual leaves). Conclusions In all, these results suggest SLEAP is a robust and versatile tool for high-throughput automated tracking of plants, presenting a new avenue for research focusing on plant dynamics.
Why it matches plant phenotyping methods植物の成長運動を抽出するため、SLEAPを植物追跡へ適応し、多様な器官・撮像条件・形態・運動で精度を評価している。植物表現型取得手法が中心である。
abstractHere we adopted the Social LEAP Estimates Animal Poses (SLEAP) framework for plant tracking.
Reproduction assets foundThe paper's Availability of data and materials statement points to a public Zenodo deposit containing the paper-specific time-lapse videos, SLEAP .slp labelled training files, and predicted output analysis files used in this study.Dataset · publicThe datasets during and/or analysed during the current study available at: https://zenodo.org/record/5764169#.YbCK0_FBxqt , https://doi.org/10.5281/zenodo.5764169 , which includes: (1) raw videos of the timelapse for each analysis. (2) The.slp files for each video analysis, which can be loaded into SLEAP and contain the 5, 10 or 20 labelled training frames.Open asset ↗zenodo · 10.5281/zenodo.5764169lines:134-177Code / dataset availability confirmedEurope PMC · OpenAlex · checked 8 Sept 2026
This work focuses on the problem of non-contact measurement for vegetables in agricultural automation. The application of computer vision in assisted agricultural production significantly improves work efficiency due to the rapid development of information technology and artificial intelligence. Based on object detection and stereo cameras, this paper proposes an intelligent method for vegetable recognition and size estimation. The method obtains colorful images and depth maps with a binocular stereo camera. Then detection networks classify four kinds of common vegetables (cucumber, eggplant, tomato and pepper) and locate six points for each object. Finally, the size of vegetables is calculated using the pixel position and depth of keypoints. Experimental results show that the proposed method can classify four kinds of common vegetables within 60 cm and accurately estimate their diameter and length. The work provides an innovative idea for solving the vegetable's non-contact measurement problems and can promote the application of computer vision in agricultural automation.
Why it matches plant phenotyping methods野菜の長さ・直径という植物器官形質を、ステレオカメラ、深度画像、キーポイント検出で非接触推定する手法が研究の中心であるため。
abstractThis work focuses on the problem of non-contact measurement for vegetables in agricultural automation.
Reproduction assets foundThe authors publicly released both the vegetable keypoint dataset (1600 COCO-format images with ROI boxes and six keypoints) and the implementation code for their size estimation method on GitHub. Labelme is a generic third-party annotation tool and is excluded.Code · publicThe implementation code of our size estimation method can be accessed on https://github.com/BourneZ130/VegetableDetection , accessed on 15 February 2022.Open asset ↗BourneZ130/VegetableDetectionlines:76-141Plant phenotyping relevance match · UnverifiedEurope PMC · checked 9 Sept 2026
Abstract Background: To investigate the effect of drought stress on wheat posture. Methods: An image acquisition system based on an infrared tracing robot was developed and a graphical user interface (GUI) software was designed to simplify the operation control of the robot. In this experiment, three genotypes of wheat, Ruihuamai 523, Jimai 22 and Xumai 33, were grown in indoor pots, and the images of wheatposture from flowering stage to maturity were collected to extract morphological parameters such as plant height, stem width and leaf inclination angle. Results: The experimental results showed that the deviation of linear trajectory was less than 3 mm when the robot traveled in a straight line at 0.4 m/s in the greenhouse; the image acquisition efficiency was about 18 images/min;the collected pictures of drought-stressed and control potted wheat groups can be used for posture assessment; the accuracy of wheat plant height and stem width with manual acquisition was 84.6% and 79.2%, respectively. After statistical analysis, it was concluded that drought stress and genotype and other influencing factors had no significant effect on plant height and stem width, but had a greater effect on leaf inclination angle. Conclusions: Therefore, this system can be used for posture collection of wheat.
Why it matches plant phenotyping methods小麦形态表型采集系统及其机器人、GUI和测量准确性验证是论文核心，而非仅将性状作为生物学实验的常规结果。
abstractAn image acquisition system based on an infrared tracing robot was developed and a graphical user interface (GUI) software was designed to simplify the operation control of the robot.
Precision management of agricultural fields as well as plant breeding are central factors for keeping yields high and to provide food, feed, and fiber for our society. A key element in breeding trials but also for targeted management actions is to analyze the growth state of individual plants objectively and at a large scale. In this letter, we address the problem of analyzing crops in real agricultural fields based on camera data recorded with mobile robots and to derive information about the plant development, e.g., to monitor phenotypic traits such as growth stage. We propose a novel single-stage object detection approach that localizes crops and weeds in the field. At the same time, it detects plant-specific leaf keypoints intending to estimate leaf count at a plant level, which is a key trait for classifying the growth stage. We implemented and thoroughly tested our approach on real sugar beet fields. As our experiments show, it performs the required detections and shows superior performance with respect to a state-of-the-art two-stage approach based on Mask R-CNN.
Why it matches plant phenotyping methods圃場ロボット画像から個体ごとの葉数を推定し、生育段階という植物形質を抽出する検出手法を開発・比較評価しており、フェノタイピング手法が中心である。
titleJoint Plant Instance Detection and Leaf Count Estimation for In-Field Plant Phenotyping
X-ray / CTPose / keypoint estimationGrowth / development / phenology
In agricultural robotics, a unique challenge exists in the automated planting of bulbous plants: the estimation of the bulb's growth direction. To date, no existing work addresses this challenge. Therefore, we propose the first robotic vision framework for the estimation of a plant bulb's growth direction. The framework takes as input three x-ray images of the bulb and extracts shape, edge, and texture features from each image. These features are then fed into a machine learning regression algorithm in order to predict the 2D projection of the bulb's growth direction. Using the x-ray system's geometry, these 2D estimates are then mapped to the 3D world coordinate space, where a filtering on the estimate's variance is used to determine whether the estimate is reliable. We applied our algorithm on 27,200 x-ray simulations from T. Apeldoorn bulbs on a standard desktop workstation. Results indicate that our machine learning framework is fast enough to meet industry standards (<0.1 seconds per bulb) while providing acceptable accuracy (e.g. error < 30° in 98.40% of cases using an artificial 3-layer neural network). The high success rates of the proposed framework indicate that it is worthwhile to proceed with the development and testing of a physical prototype of a robotic bulb planting system.
Why it matches plant phenotyping methodsX線画像と機械学習により球根の成長方向という植物器官の形態特性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。
abstractwe propose the first robotic vision framework for the estimation of a plant bulb's growth direction.
Measurement of plant nitrogen (N), phosphorus (P), and potassium (K) levels are important for determining precise fertilization management approaches for crops cultivated in greenhouses. To accurately, rapidly, stably, and nondestructively measure the NPK levels in tomato plants, a nondestructive determination method based on multispectral three-dimensional (3D) imaging was proposed. Multiview RGB-D images and multispectral images were synchronously collected, and the plant multispectral reflectance was registered to the depth coordinates according to Fourier transform principles. Based on the Kinect sensor pose estimation and self-calibration, the unified transformation of the multiview point cloud coordinate system was realized. Finally, the iterative closest point (ICP) algorithm was used for the precise registration of multiview point clouds and the reconstruction of plant multispectral 3D point cloud models. Using the normalized grayscale similarity coefficient, the degree of spectral overlap, and the Hausdorff distance set, the accuracy of the reconstructed multispectral 3D point clouds was quantitatively evaluated, the average value was 0.9116, 0.9343 and 0.41 cm, respectively. The results indicated that the multispectral reflectance could be registered to the Kinect depth coordinates accurately based on the Fourier transform principles, the reconstruction accuracy of the multispectral 3D point cloud model met the model reconstruction needs of tomato plants. Using back-propagation artificial neural network (BPANN), support vector machine regression (SVMR), and gaussian process regression (GPR) methods, determination models for the NPK contents in tomato plants based on the reflectance characteristics of plant multispectral 3D point cloud models were separately constructed. The relative error (RE) of the N content by BPANN, SVMR and GPR prediction models were 2.27%, 7.46% and 4.03%, respectively. The RE of the P content by BPANN, SVMR and GPR prediction models were 3.32%, 8.92% and 8.41%, respectively. The RE of the K content by BPANN, SVMR and GPR prediction models were 3.27%, 5.73% and 3.32%, respectively. These models provided highly efficient and accurate measurements of the NPK contents in tomato plants. The NPK contents determination performance of these models were more stable than those of single-view models.
Why it matches plant phenotyping methodsトマトのNPK含量という植物状態を、マルチスペクトル・3D画像、点群再構成、画像位置合わせ、回帰モデルで非破壊推定する方法を開発・定量評価しており、フェノタイピング手法が中心である。
abstracta nondestructive determination method based on multispectral three-dimensional (3D) imaging was proposed
Nondestructive plant growth measurement is essential for researching plant growth and health. A nondestructive measurement system to retrieve plant information includes the measurement of morphological and physiological information, but most systems use two independent measurement systems for the two types of characteristics. In this study, a highly integrated, multispectral, three-dimensional (3D) nondestructive measurement system for greenhouse tomato plants was designed. The system used a Kinect sensor, an SOC710 hyperspectral imager, an electric rotary table, and other components. A heterogeneous sensing image registration technique based on the Fourier transform was proposed, which was used to register the SOC710 multispectral reflectance in the Kinect depth image coordinate system. Furthermore, a 3D multiview RGB-D image-reconstruction method based on the pose estimation and self-calibration of the Kinect sensor was developed to reconstruct a multispectral 3D point cloud model of the tomato plant. An experiment was conducted to measure plant canopy chlorophyll and the relative chlorophyll content was measured by the soil and plant analyzer development (SPAD) measurement model based on a 3D multispectral point cloud model and a single-view point cloud model and its performance was compared and analyzed. The results revealed that the measurement model established by using the characteristic variables from the multiview point cloud model was superior to the one established using the variables from the single-view point cloud model. Therefore, the multispectral 3D reconstruction approach is able to reconstruct the plant multispectral 3D point cloud model, which optimizes the traditional two-dimensional image-based SPAD measurement method and can obtain a precise and efficient high-throughput measurement of plant chlorophyll.
Why it matches plant phenotyping methods温室トマトのクロロフィル量を推定するための統合マルチスペクトル3D計測システム、画像登録、3D再構成、推定モデルを開発・比較しており、植物表現型取得が中心である。
abstractIn this study, a highly integrated, multispectral, three-dimensional (3D) nondestructive measurement system for greenhouse tomato plants was designed.
For many robotic harvesting applications, position and angle between plant parts is required to optimally position the end-effector before attempting to approach, grasp and cut the product. A method for estimating the angle between plant parts, e.g. stem and fruit, is presented to support the optimisation of grasp pose for harvest robots. The hypothesis is that from colour images, this angle in the horizontal plane can be accurately derived under unmodified greenhouse conditions. It was hypothesised that the location of a fruit and stem could be inferred in the image plane from sparse semantic segmentations. The paper focussed on 4 sub-tasks for a sweet-pepper harvesting robot. Each task was evaluated under 3 conditions: laboratory, simplified greenhouse and unmodified greenhouse. The requirements for each task were based on the end-effector design that required a 25° positioning accuracy. In Task I, colour image segmentation for classes back-ground, fruit and stem plus wire was performed, meeting the requirement of an intersection-over-union > 0.58. In Task II, the stem pose was estimated from the segmentations. In Task III, centres of the fruit and stem were estimated from the output of previous tasks. Both centre estimations In Tasks II and III met the requirement of 25 pixel accuracy on average. In Task IV, the centres were used to estimate the angle between the fruit and stem, meeting the accuracy requirement of 25° for 73% of the cases. The work impacted on the harvest performance by increasing its success rate from 14% theoretically to 52% in practice under unmodified conditions.
Why it matches plant phenotyping methods色画像から果実と茎の位置・角度という植物器官の幾何学的形質を推定する手法を開発し、複数条件で精度検証している。単なる収穫対象の検出ではなく、器官間角度を定量化しているため対象範囲に該当する。
abstractA method for estimating the angle between plant parts, e.g. stem and fruit, is presented
Pepper / chilliLiDAR / point cloudFruitPose / keypoint estimation
The space pose of fruits is necessary for accurate detachment in automatic harvesting. This study presents a novel pose estimation method for sweet pepper detachment. In this method, the normal to the local plane at each point in the sweet-pepper point cloud was first calculated. The point cloud was separated by a number of candidate planes, and the scores of each plane were then separately calculated using the scoring strategy. The plane with the lowest score was selected as the symmetry plane of the point cloud. The symmetry axis could be finally calculated from the selected symmetry plane, and the pose of sweet pepper in the space was obtained using the symmetry axis. The performance of the proposed method was evaluated by simulated and sweet-pepper cloud dataset tests. In the simulated test, the average angle error between the calculated symmetry and real axes was approximately 6.5°. In the sweet-pepper cloud dataset test, the average error was approximately 7.4° when the peduncle was removed. When the peduncle of sweet pepper was complete, the average error was approximately 6.9°. These results suggested that the proposed method was suitable for pose estimation of sweet peppers and could be adjusted for use with other fruits and vegetables.
Why it matches plant phenotyping methods甜ピーマン果実の三次元姿勢を点群から推定する手法を開発し、シミュレーションおよび実データで誤差検証している。収穫対象の単なる検出ではなく、果実器官の姿勢という再利用可能な形態状態を定量化しているため、中心的なフェノタイピング手法に該当する。
abstractThis study presents a novel pose estimation method for sweet pepper detachment.
Systems for collecting image data in conjunction with computer vision techniques are a powerful tool for increasing the temporal resolution at which plant phenotypes can be measured non-destructively. Computational tools that are flexible and extendable are needed to address the diversity of plant phenotyping problems. We previously described the Plant Computer Vision (PlantCV) software package, which is an image processing toolkit for plant phenotyping analysis. The goal of the PlantCV project is to develop a set of modular, reusable, and repurposable tools for plant image analysis that are open-source and community-developed. Here we present the details and rationale for major developments in the second major release of PlantCV. In addition to overall improvements in the organization of the PlantCV project, new functionality includes a set of new image processing and normalization tools, support for analyzing images that include multiple plants, leaf segmentation, landmark identification tools for morphometrics, and modules for machine learning.
Why it matches plant phenotyping methods植物画像から表現型を抽出するオープンソース解析ソフトウェアの開発・改良が中心であり、植物フェノタイピング手法論文に該当する。
abstractPlant Computer Vision (PlantCV) software package, which is an image processing toolkit for plant phenotyping analysis.
Reproduction assets foundThe paper explicitly states that scripts, notebooks, SQL schema, and simple input data for its figures/results are on GitHub, that Setaria images come from publicly available datasets on the PlantCV data page, and that PlantCV v2.1 is archived on Zenodo.Code · publicScripts, notebooks, SQL schema, and simple input data associated with the figures and results presented in this paper are available on GitHub at https://github.com/danforthcenter/plantcv-v2-paper .Open asset ↗danforthcenter/plantcv-v2-paperlines:27-34Dataset · publicImages of Setaria viridis (A10) and Setaria italica (B100) are from publicly available datasets that are available at http://plantcv.danforthcenter.org/pages/data.htmlOpen asset ↗lines:27-34Plant phenotyping relevance match · UnverifiedEurope PMC · OpenAlex · checked 15 Sept 2026
Systems for collecting image data in conjunction with computer vision techniques are a powerful tool for increasing the temporal resolution at which plant phenotypes can be measured non-destructively. Computational tools that are flexible and extendable are needed to address the diversity of plant phenotyping problems. We previously described the Plant Computer Vision (PlantCV) software package, which is an image processing toolkit for plant phenotyping analysis. The goal of the PlantCV project is to develop a set of modular, reusable, and repurposable tools for plant image analysis that are open-source and community-developed. Here we present the details and rationale for major developments in the second major release of PlantCV. In addition to overall improvements in the organization of the PlantCV project, new functionality includes a set of new image processing and normalization tools, support for analyzing images that include multiple plants, leaf segmentation, landmark identification tools for morphometrics, and modules for machine learning.
Why it matches plant phenotyping methods植物フェノタイピング用の画像解析ソフトウェアを開発し、葉分割・形態計測ランドマーク・機械学習などの機能を技術的に提示しているため、方法が中心である。
abstractnew functionality includes a set of new image processing and normalization tools, support for analyzing images that include multiple plants, leaf segmentation, landmark identification tools for morphometrics, and modules for machine learning.
Systems for collecting image data in conjunction with computer vision techniques are a powerful tool for increasing the temporal resolution at which plant phenotypes can be measured non-destructively. Computational tools that are flexible and extendable are needed to address the diversity of plant phenotyping problems. We previously described the Plant Computer Vision (PlantCV) software package, which is an image processing toolkit for plant phenotyping analysis. The goal of the PlantCV project is to develop a set of modular, reusable, and repurposable tools for plant image analysis that are open-source and community-developed. Here we present the details and rationale for major developments in the second major release of PlantCV. In addition to overall improvements in the organization of the PlantCV project, new functionality includes a set of new image processing and normalization tools, support for analyzing images that include multiple plants, leaf segmentation, landmark identification tools for morphometrics, and modules for machine learning.
Why it matches plant phenotyping methods植物画像から表現型を抽出するオープンソース解析ソフトウェアの開発・機能拡張が中心であり、明確なフェノタイピング手法論文です。
abstractnew functionality includes a set of new image processing and normalization tools, support for analyzing images that include multiple plants, leaf segmentation, landmark identification tools for morphometrics, and modules for machine learning.
This paper presents a novel pose estimation algorithm for stem position detection in Japanese green pepper automatic harvesting. When the available visual cues do not provide sufficient information to the harvesting robot, information about the pose of the fruit in space is necessary for accurate stem position detection. In the proposed method the orientation of a fruit in space is obtained by fitting a model to surface points of the fruit. These surface points are acquired using a Lidar type laser range finder, and the point matching is performed using a coherent point drift algorithm with two model transformation methods, rigid and affine. The performance of the proposed method was evaluated both under laboratory conditions and in a greenhouse. In the laboratory test, the mean total error for the affine transformation was less than 25mm in 42 of 49 positions, less than 20mm in 28 of 49 positions and less than 15mm in 19 of 49 positions. For the rigid transformation, the same error was less than 25mm in 39 of 49 positions, less than 20mm in 31 of 49 positions and less than 15mm in 11 of 49 positions. The total error of the affine transformation was found to be proportional to the inclination angle, as the mean error was 11mm, 15mm, and 23mm for inclination angles of 15, 30 and 45 degrees, respectively. No relationship was found between the mean total error and the inclination angle for the rigid transformation, as the calculated mean total error was 20mm, 18mm, and 20mm for inclination angles of 15, 30 and 45 degrees, respectively. In the greenhouse test, the stem was calculated to be within the cutting range for 81 of 107 instances for affine transformation and for 66 of 107 for rigid transformation. These results suggest that the proposed method is suitable for stem position detection in the automatic harvesting of green pepper, and could be adjusted for use with other fruits and vegetables.
Why it matches plant phenotyping methodsLiDAR点群とモデル適合により果実の3D姿勢を推定する手法を開発し、実験室および温室で誤差評価している。収穫ロボット用途だが、果実の向きという再利用可能な器官形質を定量化する技術が中心である。
abstractThis paper presents a novel pose estimation algorithm for stem position detection in Japanese green pepper automatic harvesting.