UAV-based multi-view reconstruction is an important approach for high-precision, non-destructive 3D crop phenotyping. However, in greenhouse environments, UAV image acquisition is often restricted to sparse viewpoints because of UAV-induced airflow disturbances and the structural complexity of the greenhouse, which severely hinders accurate 3D phenotyping. To address this challenge, this study develops a task-driven phenotyping framework for constrained UAV viewpoints in greenhouse environments, integrating a vision-triggered flight planning strategy with an improved sparse-view 3DGS pipeline, termed SparseBerry-3DGS, for multi-view image acquisition and 3D phenotyping of greenhouse strawberries in GNSS-denied environments. Specifically, 3D Gaussian Splatting (3DGS) is improved by incorporating flow-guided initialization, depth supervision, and an adaptive pruning strategy, which effectively alleviate geometric collapse and floating artifacts under sparse-view conditions. Furthermore, sequential semantic masks generated by SAM2 are utilized to guide the segmentation of strawberry point clouds, thereby reducing background interference and segmentation errors. Experimental results show that the vision-triggered flight strategy enables stable capture of 16 surrounding images for each target fruit. Under sparse-view conditions, SparseBerry-3DGS improves reconstruction stability, with the average peak signal-to-noise ratio (PSNR) reaching 18.25 dB, corresponding to an 18% improvement. The SAM2-based segmentation module achieves high accuracy, with the mean intersection over union (mIoU) above 0.95. Geometric evaluation based on strawberry longitudinal diameter yielded an of 0.88, supporting the accuracy of fruit-scale geometric reconstruction. For weight estimation, five-fold cross-validation yielded an of 0.90 and an RMSE of 3.62 g, showing better predictive performance than models based on 2D projected area and standard 3DGS point clouds. This study provides a new approach for high-throughput, non-invasive digital crop phenotyping in greenhouse horticulture.
Why it matches plant phenotyping methods温室イチゴの3D形状再構成と重量推定を目的に、制約視点UAV撮影、SparseBerry-3DGS再構成、点群セグメンテーションを統合した表現型取得手法を開発・検証しており、方法が研究の中心である。
abstractthis study develops a task-driven phenotyping framework for constrained UAV viewpoints in greenhouse environments
Maize ear geometry (length, width, curvature, and volume) is closely tied to yield and grain-filling outcomes, but existing high-throughput phenotyping pipelines remain constrained by the cost, labor, and specialized hardware they require. We developed and validated a low-cost pipeline that reconstructs a watertight 3-D mesh of a maize ear from a single 20-second video captured with a consumer-grade DSLR on a motorized turntable under uniform LED illumination. Camera poses from a multi-seed COLMAP procedure initialize a Neural Radiance Field (NeRF), and a cylindrical holder of known diameter, visible in every frame, provides automatic metric scaling with downstream geometric quality control. Applied to 300 ears spanning a diverse maize inbred panel, 250 (83.3%) passed automated processing and quality control. Skeleton length agreed with manual caliper measurements across all 250 ears (R^2 = 0.964, RMSE = 4.68 mm), and convex-hull volume agreed with water-displacement volume on a 15-ear subset spanning the full size range (R^2 = 0.982, RMSE = 5.26 mL). Residual length error grew with ear curvature, whereas bounding-box height, which records the same straight-line chord as calipers, showed no such trend; the discrepancy therefore originates in the measurement definition, since calipers record the chord while skeleton length traces the geodesic arc. The capture hardware costs approximately 607 USD, and operator involvement fell from roughly five minutes to one minute per ear, with all downstream processing running unattended. The platform provides a foundation for breeding-scale 3-D ear phenotyping.
Why it matches plant phenotyping methodsトウモロコシ雌穂の3D形態形質を抽出する低コスト画像計測パイプラインを開発し、手動測定および体積測定で技術検証しているため、方法が研究の中心である。
abstractWe developed and validated a low-cost pipeline that reconstructs a watertight 3-D mesh of a maize ear
NeRF / 3D Gaussian SplattingFruit2D/3D reconstructionSegmentation
High-fidelity and lightweight 3D fruit models are a crucial foundation for phenotypic analysis and automated agricultural robotic operations. However, in complex agricultural scenarios characterized by varying illumination and foliage occlusion, existing 3D reconstruction methods struggle to balance reconstruction accuracy and model size, often generating massive redundant background primitives. To address this challenge, this paper proposes a novel framework for lightweight and high-fidelity 3DGS fruit reconstruction via geometric-semantic joint constraints. Specifically, the method first integrates depth priors and semantic information through a Depth-Guided Semantic Segmentation module to extract accurate target fruit masks. Next, it eliminates background noise points from the initial point cloud using a multi-view Reprojection Consistency Voting mechanism. Simultaneously, a Stochastic Background Regularized Hybrid Loss is introduced during the 3DGS training phase to decouple density and color optimization, thereby suppressing the regeneration of background Gaussian primitives. Experimental results on a multi-category fruit dataset demonstrate that while maintaining a high novel view synthesis quality (PSNR of 31.87 dB), our proposed method reduces the average model size from 230.42 MB to 62.66 MB (a 72.8% reduction), achieving robust, high-fidelity, and lightweight 3D fruit reconstruction.
Why it matches plant phenotyping methods果実の3D形状を抽出・再構成する画像ベース手法が研究の中心であり、果実形態のフェノタイピングに直接利用可能な方法を開発・評価している。
abstractHigh-fidelity and lightweight 3D fruit models are a crucial foundation for phenotypic analysis
Field / plotMultimodalNeRF / 3D Gaussian SplattingLiDAR / point cloudRGB / grayscaleStereoFruitWhole plant / canopy / plot / fieldMorphology / geometry measurementObject detection
Binocular stereo vision is a low-cost and scalable 3D perception technology that shows strong potential in agricultural phenotyping and smart agriculture. By estimating depth from multi-view RGB images, it enables non-contact, high-precision sensing of crop structure, canopy morphology, growth dynamics, and livestock traits, providing essential support for digital and intelligent agricultural production. With recent advances in deep learning-based stereo matching, multimodal sensor fusion, and 3D reconstruction, its robustness and accuracy in complex field environments have been significantly improved. This paper systematically reviews recent progress in agricultural applications of binocular stereo vision, covering system architectures, traditional and deep learning-based stereo matching methods, point cloud reconstruction techniques, and emerging supervision strategies such as 3D Gaussian splatting. It further summarizes key applications, including high-throughput phenotyping, fruit localization and robotic harvesting, weed detection and precision spraying, autonomous navigation, and livestock body condition assessment, highlighting its role in multi-task agricultural perception systems. Finally, the paper discusses major challenges, including low-texture matching difficulty, occlusions in complex environments, cross-domain generalization, real-time lightweight deployment, and limited dataset availability. Future directions are outlined in foundation model-based visual perception, self- and weakly supervised learning, multimodal fusion, and edge-efficient model design, aiming to support large-scale deployment in smart agriculture.
Why it matches plant phenotyping methods農業における双眼ステレオビジョンのシステム、ステレオマッチング、3D再構成を体系的にレビューし、作物構造・群落形態・生育動態の非接触計測とハイスループット表現型解析を主要対象としているため。
abstractThis paper systematically reviews recent progress in agricultural applications of binocular stereo vision, covering system architectures, traditional and deep learning-based stereo matching methods, point cloud reconstruction techniques
Monitoring continuous agricultural canopies is fundamentally limited by the geometric constraints and computational bottlenecks of traditional 3D reconstruction. This study presents a 3D volumetric phenotyping pipeline integrating semantic mask generation with 3D Gaussian Splatting (3DGS) to quantify greenhouse cucumber canopy architecture. To drive component-specific optimization, we evaluated custom-trained convolutional networks (YOLO11) against a zero-shot foundation model (SAM3), determining that SAM3 provided the necessary boundary precision for accurate spatial isolation. The optimized 3DGS model outperformed implicit NeRF baselines, preserving fine-scale morphological details at real-time rendering speeds ( ≈ 48 FPS). To enable actionable measurement, a uniform voxelization protocol was applied to the point cloud, successfully neutralizing algorithmic densification bias. This technical framework yielded highly accurate physical geometry, achieving a Root Mean Square Error (RMSE) of ≤ 0.59 cm against in situ leaf measurements. Transitioning to agronomic interpretation, the pipeline was deployed to quantify complex canopy architecture. It mathematically mapped structural congestion zones and provided a temporal validation of a standard pruning intervention, explicitly capturing the geometric increase in lower-canopy porosity and the upward translation of biomass. This framework provides a robust, scale-accurate tool for monitoring plant architecture and guiding dynamic canopy management.
Why it matches plant phenotyping methods植物キャノピーの3D形状を取得・定量化する画像ベースの表現型解析パイプラインを開発し、実測葉寸法で精度検証しているため、方法が研究の中心である。
abstractThis study presents a 3D volumetric phenotyping pipeline integrating semantic mask generation with 3D Gaussian Splatting (3DGS) to quantify greenhouse cucumber canopy architecture.
Finger millet is an important cereal crop widely cultivated worldwide for food and fodder. Breeding programs aim to select genotypes with desirable architectural traits to develop new varieties with higher yields. In this effort, accurate high-throughput plant phenotyping is essential for accelerating crop improvement. To overcome the time-consuming and labor-intensive process of manual measurements, this study presents a comprehensive 3D imaging pipeline that leverages neural radiance fields (NeRF), 3D gaussian splatting (3DGS), and its advanced extensions (e.g., Feature 3DGS and Gaussian Grouping) to reconstruct, segment, and analyze finger millet yield component traits using multi-view 2D images. First, multiple-view RGB images of a single finger millet plant were captured, and COLMAP was then utilized to estimate the camera poses of the images and reconstruct the sparse point cloud, followed by advanced 3D reconstruction through 3DGS and NeRF. Second, feature 3DGS and gaussian grouping models were used to generate the 3D gaussian representation of finger millet panicles. This single-process framework enabled the generation of high-fidelity 3D point clouds and semantic feature fields without the need for expensive depth sensors or manual annotations. Our results demonstrated the effectiveness of these models in capturing morphological variations across different panicle phenotypes, including compact versus open panicle architectures. In addition, the 3D point clouds of the panicles were utilized to extract structural traits for yield prediction, achieving biologically meaningful correlations with grain productivity. This work highlights the potential of 3DGS-based phenotyping pipelines as a low-cost, near real-time, photorealistic solution for trait quantification, segmentation, and yield estimation in real-world agricultural settings.
Why it matches plant phenotyping methods3D画像再構成・セグメンテーション・形質抽出を統合した植物フェノタイピング手法の開発が中心であり、収量関連形質の定量と予測まで技術的に評価している。
abstractthis study presents a comprehensive 3D imaging pipeline that leverages neural radiance fields (NeRF), 3D gaussian splatting (3DGS), and its advanced extensions (e.g., Feature 3DGS and Gaussian Grouping) to reconstruct, segment, and analyze finger millet yield component traits using multi-view 2D images.
SoybeanNeRF / 3D Gaussian SplattingRGB / grayscaleFruitSegmentation
Neural Radiance Fields (NeRF) have been widely adopted for reconstructing high-quality 3D scenes from 2D RGB images. However, achieving accurate 3D object segmentation within these reconstructed scenes remains challenging. Existing NeRF-based segmentation methods either rely on post-processing (SA3D), which produces noisy point clouds due to the absence of density field optimization, or employ joint training with additional segmentation heads (FruitNeRF), which can lead to suboptimal performance due to conflicting learning objectives. In this work, we propose InvNeRF-Seg (Input-substitution NeRF for Segmentation), a two-stage fine-tuning strategy for 3D object segmentation that preserves the original NeRF architecture and loss function entirely. We first train a standard NeRF on RGB images and then fine-tune it using 2D segmentation masks formatted as RGB-like inputs, without introducing any architectural modifications or additional loss functions. This input-substitution approach reshapes the density field to align with object regions while suppressing background density. We validate InvNeRF-Seg through comprehensive ablation studies examining the roles of density and color MLPs, loss function choices, and training strategies. Field density analysis reveals consistent semantic refinement: densities of object regions increase while background densities are suppressed. Experiments on synthetic fruit datasets and real-world soybean imagery demonstrate that InvNeRF-Seg produces cleaner 3D segmented point clouds compared to both SA3D and FruitNeRF, enabling more accurate downstream object counting. The method is further validated on a self-collected soybean dataset to demonstrate its applicability in real-world agricultural scenarios. Our code is available at https://github.com/ZJiangsan/InvNeRF-Seg .
Why it matches plant phenotyping methods植物画像から3D物体領域を抽出するNeRFベース手法の開発・比較検証が中心で、果実・ダイズ画像を対象に物体カウントへ応用しているため、植物器官の形態・数量推定に関わるフェノタイピング手法として含める。
abstractIn this work, we propose InvNeRF-Seg (Input-substitution NeRF for Segmentation), a two-stage fine-tuning strategy for 3D object segmentation
MilletNeRF / 3D Gaussian SplattingWhole plant / canopy / plot / fieldGrowth / time-series analysisVisualization / data managementGrowth / development / phenology
Finger millet ( Eleusine coracana (L.) Gaertn.) is a small seeded cereal known for its health benefits and its ability to grow in water-limited and saline environments. Despite its health and adaptation advantages, there is no standardized BBCH scale developed for finger millet, precluding standardization of a defined growth and development scale. The initial step in any breeding program and related research, is the precise, consistent and standardized description of crop growth stages. Therefore, we present a standardized phenological scale to describe and compare different stages of growth and development. In this study, we used four-finger millet accessions from the U.S.D.A. National Plant Germplasm System (NPGS) collection. Using a standardized BBCH chart as the baseline, we describe nine main stages of development to provide a simplified and user-friendly phenological growth staging scale for finger millet. A novel feature of this study was the use of Neural Radiance Fields (NeRF) to generate three-dimensional (3-D) renderings that align growth stages with 3-D imaging, paving the way for future phenological staging studies to incorporate 3-D visualization as a tool for standardization, researcher training, and reduction of observer bias across environments.
Why it matches plant phenotyping methods指のキビの生育段階を標準化する手法を開発し、NeRFによる3D画像化を生育ステージ判定・標準化に組み込んでいるため、植物フェノタイピング手法が中心である。
titlePhenological growth stages of Finger millet (Eleusine coracana (L.)) using 3-D phenotyping with relevance to breeding applications
Published23 Jul 2026The international archives of the photogrammetry, remote sensing and spatial information sciences/International archives of the photogrammetry, remote sensing and spatial information sciencesCited by 0 · OpenAlex ↗
Field / plotNeRF / 3D Gaussian SplattingThermalWhole plant / canopy / plot / field2D/3D reconstructionVisualization / data managementPlant / canopy temperature
Abstract. Urban trees provide critical ecosystem services in dense city environments, yet current workflows for monitoring their thermal behaviour remain confined to 2D desktop-based analysis with no three-dimensional spatial context or field-deployable visualization capability. This paper presents a complete pipeline for in-situ 3D thermal mesh visualization of urban trees in Augmented Reality (AR), combining Thermal InfraRed (TIR) image acquisition, Gaussian Splatting-based mesh reconstruction, quantitative validation, and mobile AR deployment. TIR images of a Tilia tomentosa acquired with a FLIR T560 camera are preprocessed with a standardized false-colour palette and fed into the MILo (Mesh-In-the-Loop Gaussian Splatting) framework to reconstruct a thermally attributed 3D mesh. Geometric evaluation against a Z+F IMAGER 5016 TLS reference using the M3C2 algorithm demonstrates that MILo recovers 13.5 times more canopy geometry than traditional multi-view stereo under thermal imagery, with a standard deviation of 4.0 cm. A colourmap inversion procedure recovers per-vertex temperature estimates from the GS-derived mesh colours, yielding a mean absolute difference of 0.7°C against direct T-Cam measurements (thermal camera mounted on the laser scanner), within the combined instrument accuracy of both sensors. The resulting thermal Gaussian Splat was deployed in a custom Android AR application supporting hybrid marker-based and GPS-based spatial anchoring for in-situ visualization. These results demonstrate the technical feasibility of GS-based thermal reconstruction and mobile AR as a medium for communicating three-dimensional canopy thermal information to educators and urban forestry practitioners.
Why it matches plant phenotyping methods都市樹木の葉冠温度と3D形状を取得・可視化する熱画像ベースの再構成パイプラインを開発し、TLSおよび熱カメラとの定量検証まで行っており、植物フェノタイピング手法が研究の中心である。
abstractThis paper presents a complete pipeline for in-situ 3D thermal mesh visualization of urban trees in Augmented Reality (AR), combining Thermal InfraRed (TIR) image acquisition, Gaussian Splatting-based mesh reconstruction, quantitative validation, and mobile AR deployment.
Plant phenotyping relevance match · UnverifiedOpenAlex · Crossref · Europe PMC · checked 5 Sept 2026
Introduction: Plant phenotyping requires accurate and repeatable three-dimensional structural information, but practical acquisition conditions in greenhouses, seedling rooms, and indoor pot experiments often include complex backgrounds, handheld motion blur, and thin leaf structures. These factors reduce the robustness of conventional three-dimensional reconstruction methods and limit their use in low-cost and automated phenotyping. Methods: To address this problem, this paper proposes F2DMAS, an automated three-dimensional plant phenotyping workflow using consumer-grade smartphone videos. The workflow first converts multiview RGB videos into image sequences and removes motion-blurred frames through frequency-domain quality filtering. A frequency-spatial plant segmentation module, termed FSAM3, is then introduced to separate plant structures from complex backgrounds without task-specific annotated training data. The segmented image sequences are further reconstructed using 2D Gaussian Splatting, followed by TSDF-based meshing, scale recovery, and virtual measurement for extracting plant height, canopy width, leaf length, and leaf width. Results: Experiments were conducted on 15 plant species under two acquisition scenarios. The proposed workflow achieved stable plant reconstruction under non-ideal background conditions, with PSNR, SSIM, and LPIPS values of 31.09, 0.9711, and 0.0365, respectively. Compared with the baseline reconstruction workflow, F2DMAS substantially reduced the processing time for mesh extraction while improving reconstruction quality. The extracted phenotypic traits showed strong agreement with manual measurements, with R² values ranging from 0.90 to 0.99, RMSE values ranging from 0.64 to 1.21 cm, and MAPE values ranging from 4.50% to 9.73%. Discussion: These results indicate that F2DMAS can provide an end-to-end workflow from smartphone video acquisition and plant segmentation to three-dimensional reconstruction and phenotypic trait extraction. The proposed method offers a practical and deployable solution for greenhouse seedling cultivation, potted plant experiments, and low-cost three-dimensional plant phenotyping.
Why it matches plant phenotyping methodsスマートフォン動画から植物の3D構造を再構成し、複数の形態形質を抽出・検証するワークフロー自体が中心的な方法論的貢献である。
abstractThe extracted phenotypic traits showed strong agreement with manual measurements, with R² values ranging from 0.90 to 0.99
Practical three-dimensional (3D) phenotyping in large-scale orchards with repetitive row structures remains challenging, and systematic evidence comparing both accuracy and acquisition efficiency under outdoor conditions remains limited. This study presents a field-deployable evaluation framework and implements it in a 2-ha commercial Japanese pear orchard trained under a joint V-trellis system. Using a terrestrial laser scanner (TLS) as the reference, we evaluated two handheld LiDAR systems (a low-cost SLAM-based system and a high-performance system), structure from motion / multi-view stereo (SfM/MVS) reconstructions from three camera platforms (a digital camera, an action camera, and a 360° camera), and 3D Gaussian splatting (3DGS) constructed from action-camera video. Measurements were taken at two spatial scales to capture scale-dependent effects. In the span-scale survey (4 m), location error was derived from TLS-referenced target coordinate differences, and reconstruction error was quantified using cloud-to-mesh distances with cubic targets. In the row-scale survey (one tree row), positional stability during continuous mapping was evaluated as location error. Operational metrics (acquisition time, data volume, and processing effort) were also documented. The results demonstrate clear trade-offs among the methods: LiDAR enables rapid wide-area acquisition but is susceptible to cumulative drift in row-structured environments, whereas SfM/MVS provides superior geometric fidelity at the cost of increased time and data volume. Although 3DGS is less suitable for precise quantitative measurement, it demonstrates strong potential for intuitive visualization of orchard structure and fruit distribution. These findings highlight the need for staged, purpose-specific, and seasonally adaptive strategies for orchard-scale digital twin development.
Why it matches plant phenotyping methods果樹園における複数の3D取得法を比較・評価し、樹体構造や果実分布の定量的取得精度、安定性、運用性を検証しており、植物フェノタイピング手法が中心である。
abstractPractical three-dimensional (3D) phenotyping in large-scale orchards with repetitive row structures remains challenging
NeRF / 3D Gaussian SplattingLiDAR / point cloudLeafWhole plant / canopy / plot / field2D/3D reconstructionGrowth / time-series analysisTrackingArchitecture / morphology / geometryGrowth / development / phenology
Quantifying plant growth dynamics from sparse longitudinal 3D observations is fundamental for agriculture and plant sciences. Yet, plants pose unique challenges: they undergo intricate non-rigid deformations, exhibit changing topology as new organs emerge, and often lack explicit temporal correspondences between consecutive data acquisitions due to newly formed tissue. Methods designed for general scenes struggle to model topology changes and asynchronous organ growth characteristic of plants. To address these challenges, we introduce GrowFields, a compositional dynamic neural field representation for organ-aware 4D plant growth modelling from point cloud time series. Our approach decomposes a plant into its constituent organs and aligns each organ into its own canonical coordinate frame, isolating intrinsic growth patterns from global plant motion. We then learn a shared continuous neural deformation field that models temporal dynamics across all organs, conditioned on learnable per-organ latent codes capturing organ identity and growth characteristics. The resulting modular yet unified representation naturally accommodates the asynchronous development of plant organs while remaining grounded in the practical setting of organ-level plant tracking. We evaluate GrowFields on growth sequences from four plant species, assessing geometric fitting and organ tracking accuracy using manually annotated leaf-tip trajectories. Results demonstrate consistent improvements in spatial precision, temporal coherence, and morphological fidelity over a range of existing representations.
Why it matches plant phenotyping methods植物の3D時系列観測から器官追跡と成長動態を抽出する、オルガン認識型4Dニューラルフィールド手法の開発・評価が中心である。
abstractwe introduce GrowFields, a compositional dynamic neural field representation for organ-aware 4D plant growth modelling from point cloud time series.
Abstract. 3D instance segmentation for laser scanning (LiDAR) point clouds remains a challenge in many remote sensing-related domains. Successful solutions typically rely on supervised deep learning and manual annotations, and consequently focus on objects that can be well delineated through visual inspection and manual labeling of point clouds. However, for tasks with more complex and cluttered scenes, such as in-field plant phenotyping in agriculture, such approaches are often infeasible. In this study, we tackle the task of in-field wheat head instance segmentation directly from terrestrial laser scanning (TLS) point clouds. To address the problem and circumvent the need for manual annotations, we propose a novel two-stage pipeline. To obtain the initial 3D instance proposals, the first stage uses 3D-to-2D multi-view projections, the Grounded SAM pipeline for zero-shot 2D object-centric segmentation, and multi-view label fusion. The second stage uses these initial proposals as noisy pseudo-labels to train a supervised 3D panoptic-style segmentation neural network. Our results demonstrate the feasibility of the proposed approach and show performance improvements relative to Wheat3DGS, a recent alternative solution for in-field wheat head instance segmentation without manual 3D annotations based on multi-view RGB images and 3D Gaussian Splatting, showcasing TLS as a competitive sensing alternative. Moreover, the results show that both stages of the proposed pipeline can deliver usable 3D instance segmentation without manual annotations, indicating promising, low-effort transferability to other comparable TLS-based point cloud segmentation tasks.
Why it matches plant phenotyping methods圃場小麦穂の3D個体セグメンテーションを対象に、TLS点群とゼロショット・マルチビュー融合・擬似ラベル学習を組み合わせた表現型取得手法を開発・評価しており、方法が研究の中心である。
abstractsuch as in-field plant phenotyping in agriculture, such approaches are often infeasible.
Computed tomography enables non-destructive phenotyping of fruit internal structure but traditionally requires hundreds of projections, limiting throughput. Under sparse-view conditions, conventional and learning-based methods both suffer from streaking artifacts and regional distortions that degrade trait quantification. This study present CitrusGS, an integrated framework that achieves high-fidelity 3D reconstruction and precise morphological phenotyping of citrus fruit from only 15 projections using radiative 3D Gaussian splatting. Our method employs sparse-point initialization, optimized loss composite, and dual-stage pruning to suppress artifacts while preserving anatomically critical details with significantly higer convergence efficiency. In the citrus fruit datasets, CitrusGS achieves 29.78 dB PSNR and 0.870 SSIM, outperforming corresponding baseline method by 1.58 dB and 0.067 in SSIM, and enables automated extraction of ten external and internal phenotypic traits with R 2 larger than 0.944. Moreover, the framework shows initial zero-shot transferability across pathological citrus samples and additional horticultural specimens without retraining. By reconciling acquisition efficiency with anatomical fidelity using low-cost X-ray hardware, CitrusGS provides a promising framework for high-throughput, non-destructive phenotyping in breeding and grading applications.
Why it matches plant phenotyping methods柑橘果実の疎視野CT再構成法を開発・検証し、内部・外部形質を自動抽出するフェノタイピングが中心である。
abstractThis study present CitrusGS, an integrated framework that achieves high-fidelity 3D reconstruction and precise morphological phenotyping of citrus fruit from only 15 projections using radiative 3D Gaussian splatting.
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 · publicOur codes are available at https://github.com/Petrichoror/CitrusGS .Open asset ↗Petrichoror/CitrusGSlines:325-387Plant phenotyping relevance match · UnverifiedOpenAlex · arXiv · checked 5 Sept 2026
3D plant phenotyping is notoriously known to be procedure-complicated and of low throughput due to the extensive multi-view imaging, the fragile 3D reconstruction pipeline, and the additional cost from reconstructed geometry to phenotypic extraction. These limitations are further amplified in low-cost data acquisition, where smartphone videos or sparsely sampled multi-view images provide limited view overlap and self-occlusion. In this work, we show that the conventional 3D plant phenotyping pipeline could be streamlined and significantly accelerated with 3D Foundation Models (3DFMs), and particularly, present one of the first cross-crop 3D phenotyping frameworks powered by 3DFMs. The framework replaces COLMAP-style sparse initialization with 3DFM-based feed-forward geometric recovery, combines geometry-constrained 3D Gaussian Splatting for dense reconstruction, enables few-view reconstruction through iterative view synthesis and refinement, and converts reconstructed geometry into measurable organs through 2D-to-3D semantic transfer, metric scale recovery, and organ instance separation. We further construct a cross-crop dataset with smartphone-based image acquisition, diverse plant morphologies, and manual annotations for segmentation and phenotypic evaluation. Experiments across 26 plant sequences show that 3D Foundation Models reduce the average reconstruction time from 6.52 minutes to 1.58 seconds while maintaining high reconstruction quality and phenotyping accuracy. These results suggest a fresh technical route for high-throughput 3D plant phenotyping, from low-cost image acquisition to fast reconstruction, perception, scale recovery, and phenotypic measurement.
Why it matches plant phenotyping methods3D画像再構成から器官分離・形質測定までを統合した高速植物フェノタイピング手法を開発し、複数作物・データセットで性能評価しているため。
abstractpresent one of the first cross-crop 3D phenotyping frameworks powered by 3DFMs
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 · UnverifiedOpenAlex · Europe PMC · Crossref · checked 5 Sept 2026
Accurate characterization of plant 3D architecture and semantic parsing of key cotton organs represent an essential prerequisite for crop phenotyping, precision field management, and cultivar breeding and selection. However, the overall 3D structure of field-grown cotton plants is highly complex in field environments. The morphological traits and spatial distribution of various organs are modulated by multiple factors including genetics, environmental conditions, and cultivation management practices, resulting in pronounced phenotypic variation under field conditions. Thus, this study proposed an integrated technical framework for UAV-based 3D reconstruction and organ semantic segmentation tailored for field-grown cotton. High-fidelity 3D point clouds of the crop canopy are generated by coupling an field-adapted low-altitude unmanned aerial vehicle (UAV) acquisition strategy with neural radiance fields (NeRF). Despite these attractive characteristics, segmenting and analyzing such intricate point clouds can be quite challenging. To effectively parse these complex geometric structures, a novel deep learning architecture named FieldCotSeg-Net is introduced. This model integrates an Anisotropy-aware Local Attention (ALA) module and a Hierarchical Feature Refinement Gate (HFRG) module to capture fine-grained features for precise point cloud segmentation. Experimental results demonstrate that the proposed model achieves outstanding performance on both the Huaxing3Dcot and Crops3D cotton datasets. On the Huaxing3Dcot dataset, the model yields a mean intersection over union (mIoU) of 75.7%, representing a 4.5% improvement over the baseline model. On the Crops3D cotton dataset, after retraining the model on this dataset, it attains an mIoU of 74.5%, showing substantial adaptability and effectiveness in organ segmentation. This technical system provides a scalable and robust solution for field cotton phenotyping analysis in breeding research and commercial production.
Why it matches plant phenotyping methods綿花の圃場表現型解析を目的に、UAVによる3D再構成と器官セグメンテーション手法を開発・評価しており、植物形態の取得が研究の中心である。
abstractThis technical system provides a scalable and robust solution for field cotton phenotyping analysis in breeding research and commercial production.
Three-dimensional (3D) reconstruction of seedlings commonly suffers from under-constrained issues in areas with severe occlusion and insufficient observation, leading to false geometry and rendering artifacts. The existing methods primarily rely on pixel-level consistency constraints, resulting in poor stability when reconstructing seedlings with slender branches and complex structures. To address these challenges, this paper proposes a 3D reconstruction method for seedlings which is based on improved DIFIX3D+. By integrating the Stable Diffusion-VAE module into the DIFIX3D+ framework, this approach enhances the reconstruction constraint mechanism at the representation level. Firstly, high-precision segmentation of seedlings was achieved by eliminating background interference using the Florence-2 semantic priors and the SAM2 (Segment Anything Model 2) instance segmentation model. Then, by jointly mapping DIFIX3D+ restored images and their corresponding rendered images into the latent space of Stable Diffusion-VAE, consistency supervision was established within the latent representation domain. It worked in tandem with traditional pixel-level constraints, thereby suppressing the propagation of pixel noise caused by lighting variations, specular reflections, and texture repetitions at the representation level. Experimental results demonstrated that on the self-built dataset, compared to the original 3DGS (3D Gaussian splatting) and DIFIX3D+, the proposed method achieved the peak signal-to-noise ratio (PSNR) improvement of 28.99% and 24.25%, respectively, and the structural similarity index measure (SSIM) improvement of 3.16% and 3.74%, while reducing the learned perceptual image patch similarity (LPIPS) by 9.24% and 41.94%. The average PSNR, SSIM, and LPIPS values reached 31.54 dB, 0.9759, and 0.054, respectively, demonstrating significantly superior reconstruction accuracy compared to mainstream models such as Nerfacto, Plenoxels, and Mip-Splatting. This method combines high visual fidelity with excellent geometric structure representation capabilities, providing a technical reference for high-quality, low-cost 3D reconstruction of seedlings in complex scenes.
Why it matches plant phenotyping methods苗の3D形状・構造を抽出する画像ベース再構成手法の開発が中心であり、植物表現型取得の方法論に該当する。
abstractThis method combines high visual fidelity with excellent geometric structure representation capabilities, providing a technical reference for high-quality, low-cost 3D reconstruction of seedlings in complex scenes.
The objective of this study is to propose a digital fruit tree reconstruction method based on neural radiation field theory that will enable the efficient, accurate, and non-destructive acquisition of phenotypic information from fruit trees, while simultaneously reducing the cost of collecting this data. Firstly, a low-cost information acquisition platform is constructed for the purpose of shooting a multi-view video around a fruit tree. The video is then extracted and framed using a motion recovery structural algorithm, thereby obtaining a multi-view image sequence of the fruit tree with positional data. Secondly, the the image sequence is employed to train the neural radiation field, thereby obtaining a converged three-dimensional scene of the fruit tree. Ultimately, the three-dimensional scene is derived in the form of a point cloud, thus yielding a high-phenotypic detail point cloud model of the fruit tree. A multi-period point cloud model of an apple tree in an orchard environment, encompassing the flowering, fruiting, and dormant periods, was reconstructed using the aforementioned method. The experimental results demonstrated that the error associated with the tree shape data recorded by the multi-period point cloud model established by the aforementioned method was, on average, below 5%. Furthermore, the average error across all periods was 2.69%, representing a 75.50% reduction compared to traditional reconstruction methods. The dimensional accuracy of the point cloud model at the organ scale can reach the millimetre level, with an average error of 3.10% for fruit diameter, which is 66.19% lower than that of the traditional reconstruction method. The reconstruction method is robust to all periods of fruit trees and can meet the majority of cases of digital fruit tree reconstruction.
Why it matches plant phenotyping methods果樹の多視点画像からNeRFと点群を用いて樹形・器官寸法などの表現型を非破壊取得する手法を開発・検証しており、方法が研究の中心である。
abstractpropose a digital fruit tree reconstruction method based on neural radiation field theory that will enable the efficient, accurate, and non-destructive acquisition of phenotypic information from fruit trees
Plant phenotyping relevance match · UnverifiedOpenAlex · Europe PMC · Crossref · checked 6 Sept 2026
Abstract Automated three-dimensional reconstruction of plant architecture under- pins high-throughput crop phenotyping, yet its deployment in practical field settings is constrained by two fundamental limitations of Neural Radiance Fields (NeRF): degraded geometry under sparse, unstructured image capture, and a complete absence of calibrated uncertainty estimates that would allow practitioners to distinguish reliable geometry from reconstruction artefacts. We present UA-PlantNeRF, a unified framework that resolves both limita- tions through three tightly coupled contributions. First, building on VISAR, our prior intelligent video frame-selection strategy (weights α1 = 0.4, α2 = 0.3, α3 = 0.3), the pipeline identifies maximally informative, non-redundant view- points from raw footage with as few as 15 frames. Second, a dual-head NeRF architecture augmented with Monte Carlo Dropout produces jointly decom- posed aleatoric and epistemic uncertainty alongside each reconstructed voxel, trained under a heteroscedastic negative log-likelihood objective. Third, split conformal prediction—with calibration performed on held-out plants to preserve exchangeability— yields provable, distribution-free per-ray cov- erage guarantees at any user-specified confidence level. Evaluated across three benchmarks (Pheno4D, ROSE-X, and our novel UA-Field greenhouse dataset) at sparsity levels N ∈ {15, 30, 50}, UA-PlantNeRF achieves PSNR 28.7±0.6 dB and SSIM 0.89±0.01 at N = 30, outperforming all sparse-view baselines (p
Why it matches plant phenotyping methods植物の3D構造・アーキテクチャを推定する画像ベースのNeRF手法を開発し、複数ベンチマークで評価しているため、植物表現型計測法が中心です。
abstractEvaluated across three benchmarks (Pheno4D, ROSE-X, and our novel UA-Field greenhouse dataset)
Plant phenotyping relevance match · UnverifiedOpenAlex · Europe PMC · checked 15 Sept 2026
High-fidelity 3D reconstruction and precise phenotypic parameter extraction of banana plants are critical for crop growth monitoring and yield estimation in precision agriculture. However, traditional methods encounter significant bottlenecks: LiDAR systems are cost-prohibitive for widespread adoption, while traditional photogrammetry often fails to handle the complex canopy structures, severe occlusions, and weak texture features characteristic of banana leaves. To address these limitations, this article proposes a novel framework for 3D reconstruction and automatic phenotyping based on multi-view images captured by mobile phones. We introduce BN-NeRF, an enhanced Neural Radiance Field method built upon Instant-NGP. Specifically, we integrate three key technical improvements: (1) frame-level geometric calibration to correct camera pose drift caused by handheld motion; (2) sparse geometric anchoring to explicitly constrain depth and scale using sparse point clouds; and (3) thin-leaf prior regularization to suppress artifacts and improve the geometric accuracy of leaf surfaces. Building on this reconstruction, we establish a complete pipeline to recover explicit metric geometry from implicit radiance fields. By combining mesh topological analysis with geodesic algorithms, we achieve automated and precise extraction of key morphological parameters. Extensive experiments were conducted on a dataset of 90 banana plants in a real-world orchard. The results demonstrate that BN-NeRF achieves superior rendering quality (PSNR of 32.4 dB, SSIM of 0.951, and LPIPS of 0.152) while maintaining inference speeds comparable to Instant-NGP. Furthermore, the extracted phenotypic parameters showed strong agreement with manual ground truth across both leaf-level and structural traits. In addition to trait-specific regression performance, the evaluation also includes normalized completeness analysis, calibration-cube-based scale validation, and Bland-Altman agreement analysis, supporting the measurement reliability of BN-NeRF for field phenotyping. This study demonstrates that low-cost smartphone-based acquisition, combined with BN-NeRF, can support accurate field phenotyping of banana plants. In addition, an implemented mobile-cloud system was functionally validated through repeated end-to-end runs on an iPhone 13 client and a cloud workstation.
Why it matches plant phenotyping methodsスマートフォン画像からの3D再構成と植物形態形質抽出を中核とするBN-NeRF手法を開発し、圃場データで精度・再現性を検証しているため。
abstractthis article proposes a novel framework for 3D reconstruction and automatic phenotyping based on multi-view images captured by mobile phones
Field / plotNeRF / 3D Gaussian SplattingPhotogrammetry / SfM / MVSLiDAR / point cloudStem / branchMorphology / geometry measurement2D/3D reconstructionArchitecture / morphology / geometry
Accurate characterization of tree stem geometry is essential for forest inventories, yet conventional field measurements of diameter at breast height (DBH) are limited to a single cross-section and do not capture vertical variability along the trunk. This study compares five approaches for stem characterization in a Mediterranean forest: mobile laser scanning (MLS), consumer-grade iPad-LiDAR, Structure from Motion (SfM) photogrammetry, Gaussian Splatting (GS), and manual field measurements. Data were acquired simultaneously within a 2.5 m radial plot. DBH was estimated through RANSAC-based circular fitting, and stem sections were extracted every 20 cm to assess diameter stability along the trunk. All techniques produced similar mean DBH values closely matching field measurements (23 cm), with MLS achieving the lowest RMSE (1.29 cm), followed by SfM (1.52 cm), GS (1.60 cm), and iPad-LiDAR (2.26 cm). However, marked differences were observed in vertical completeness. MLS captured the full vertical profile of the stems, reaching 14.11 m, whereas SfM and GS from iPhone, and iPad-LiDAR were limited to approximately 6 m or less. The results indicate that although low-cost image-based approaches can provide accurate DBH estimates under controlled conditions, MLS remains the most robust solution for comprehensive vertical stem characterization.
Why it matches plant phenotyping methods森林樹幹のDBHと垂直方向の形状を、複数の3Dセンシング手法で推定・比較し、RMSEや垂直完全性を評価している。植物形状計測法の技術比較・検証が中心である。
abstractThis study compares five approaches for stem characterization in a Mediterranean forest: mobile laser scanning (MLS), consumer-grade iPad-LiDAR, Structure from Motion (SfM) photogrammetry, Gaussian Splatting (GS), and manual field measurements.
High-quality, real-time three-dimensional (3D) reconstruction and phenotyping of field crops are essential for advancing digital farm management. However, challenges such as diverse crop postures and fluctuating illumination in open-field environments significantly hinder reconstruction accuracy. To address these issues, this study proposes AgriGaussian, a novel 3D phenotyping method based on 3D Gaussian Splatting (3DGS), designed to enable low-cost, high-fidelity reconstruction and phenotypic analysis of agricultural fields, with chili pepper plants as an example. Specifically, a dataset encompassing three distinct growth stages of chili pepper plants was collected and preprocessed using a phenotyping robot (PR) equipped with oblique-view imaging. The AgriGaussian method integrates three core components: (1) a Depth-Supervised Strategy (DSS) for adaptively enhancing phenotypic features based on gradient variations; (2) a Crop Appearance Extraction (CAE) module that suppresses the effects of illumination and environmental interference while preserving fine phenotypic details; and (3) a Multi-scale Chunking Training (MCT) strategy that enables scalable reconstruction of large agricultural scenes. Experimental results demonstrate that AgriGaussian achieves high-fidelity reconstruction, with peak signal-to-noise ratio (PSNR) of 26.85 dB, structural similarity index (SSIM) of 0.86, and learned perceptual image patch similarity (LPIPS) of 0.24—surpassing existing baseline algorithms. Compared to manual measurements, the reconstructed canopy height and volume exhibit strong agreement, with coefficients of determination (R2) reaching 0.90 and 0.86, respectively, and average absolute percentage errors as low as 5.1%. Furthermore, the method accurately reconstructs fine-grained features such as foliar lesions and surface textures. In summary, AgriGaussian enables large-scale, high-fidelity reconstruction of field crop phenotypes and provides a promising foundation for advancing 3D phenotyping and digital twin technologies in precision agriculture.
Why it matches plant phenotyping methods植物の3D形態・構造を抽出する低コスト画像ベース表現型解析法を開発し、手動測定との一致性も検証しているため、表現型取得法が研究の中心である。
abstractthis study proposes AgriGaussian, a novel 3D phenotyping method based on 3D Gaussian Splatting (3DGS)
Plant phenotyping relevance match · UnverifiedCrossref · OpenAlex · Europe PMC · checked 15 Sept 2026
Accurate acquisition of plant phenotypes is crucial for elucidating plant growth and development, underlying genetic mechanisms, and responses to environmental stimuli. Traditional three-dimensional (3D) phenotyping mainly captures geometric traits such as height, leaf area, and canopy volume, while overlooking physiological and biochemical information. Here, we present a hyperspectral point clouds generation method based on PlantGaussian (a 3D Gaussian Splatting technique) that integrates structural and spectral information, extending 3D phenotyping beyond geometry to include physiology. High-quality plant point clouds were first reconstructed using PlantGaussian, and hyperspectral images(HSI) were mapped onto them to produce hyperspectral point clouds. In potted soybean experiments, we built predictive models linking hyperspectral reflectance to SPAD (chlorophyll content) and EWT (equivalent water thickness), and visualized their 3D distributions. The hyperspectral point clouds achieved strong predictive performance for SPAD ( R 2 = 0.78, RMSE = 2.05) and EWT ( R 2 = 0.80, RMSE = 1.07), thereby validating the approach. It further revealed clear vertical stratification within the canopy, highlighting significant spatial heterogeneity of SPAD and EWT in individual plants. Temporal monitoring from August 6 to 21, 2025, captured a sharp increase in EWT after heavy rainfall on August 11. Overall, our results demonstrate that hyperspectral point clouds enable accurate, non-destructive trait estimation and provide a powerful tool for exploring plant function, monitoring stress responses, and advancing precision agriculture.
Why it matches plant phenotyping methods植物の3D形態とハイパースペクトル情報を統合してSPAD・EWTを推定する手法を開発し、予測性能を検証しているため、植物フェノタイピング手法が中心である。
abstractHere, we present a hyperspectral point clouds generation method based on PlantGaussian (a 3D Gaussian Splatting technique) that integrates structural and spectral information, extending 3D phenotyping beyond geometry to include physiology.
Forest digital twins play a crucial role in modern precision forestry by supporting biomass estimation and carbon cycle monitoring. However, existing 3D reconstruction methods struggle to simultaneously achieve metric-level structural accuracy and visual realism in complex understory environments. This study proposes a semantically constrained 3D Gaussian Splatting framework that fuses handheld LiDAR point clouds with unmanned aerial vehicle imagery. First, a multi-modal fusion mechanism is constructed to extract geometric anchors from registered LiDAR data for precise 3DGS spatial initialization, which mitigates rendering artifacts and geometric drift caused by poor initialization in purely visual methods. Second, a semantic regularization optimization strategy is proposed to realize differentiated modeling of tree trunks and canopies, effectively balancing the structural accuracy of rigid trunks and the photorealistic rendering of non-rigid canopies. Experiments conducted on three study plots demonstrate that the proposed approach achieves an average PSNR of 24.94 dB, SSIM of 0.773, and LPIPS of 0.231 across all plots, outperforming standard NeRF and baseline 3DGS, while enabling DBH estimation with R2 = 0.848 and RMSE = 2.705 cm. This method provides a solution for high-fidelity forest digital twin construction in open-canopy forest environments such as urban and campus forests.
Why it matches plant phenotyping methodsLiDAR・UAV画像を統合した3D再構成法を開発し、樹幹・樹冠の構造モデル化とDBH推定を評価しており、植物形質取得が中心的な技術貢献である。
abstractThis study proposes a semantically constrained 3D Gaussian Splatting framework that fuses handheld LiDAR point clouds with unmanned aerial vehicle imagery.
High-precision plant phenotyping requires efficient 3D reconstruction methods with high geometric quality. 3D Gaussian Splatting (3DGS) has recently emerged as a promising approach for real-time 3D reconstruction, achieving impressive visual quality. However, in crop environments dominated by monochromatic and low-texture regions, existing 3DGS methods often produce ambiguous geometries and fail to recover geometry-consistent 3D surfaces. To address these limitations, we propose LV-3DGS (Leafy Vegetables-3DGS), an optimized 3DGS-based framework tailored for the reconstruction of leafy vegetable scenes. First, a blurred reconstruction module is introduced to mitigate reconstruction artifacts caused by camera motion blur during multi-view image acquisition. Second, we propose a planar optimization strategy and design both local and global geometric consistency regularizations to optimize the model, thereby improving the surface reconstruction quality and geometric accuracy. Third, based on an analysis of individual Gaussian contributions, a contribution-based pruning strategy is developed to selectively remove inaccurate geometric components, achieving accurate scene geometry while reducing memory consumption and improving rendering efficiency. In addition, a quantitative geometric evaluation method is proposed for assessing reconstruction quality. Experimental results demonstrate that the proposed method achieves the highest accuracy among the tested baselines, with SSIM, PSNR, and LPIPS reaching 0.94, 34.53 dB, and 0.11, respectively. Moreover, the geometric consistency (GC) metric attains 0.317 cm. Finally, phenotypic parameters are measured from the reconstructed leafy vegetable point clouds. Compared with ground truth measurements, the proposed approach yields coefficients of determination (R2) of 0.9959, 0.9651, and 0.9895 for plant height, leaf number, and leaf area, respectively. These results are significantly outperform to some existing phenotyping methods, providing a new methodology and technical solution for high-precision, low-cost, and high-throughput crop phenotyping.
Why it matches plant phenotyping methods葉菜類の3D再構成とそこからの形質抽出を中心に、手法開発・幾何評価・実測値との検証を行っているため、植物フェノタイピング手法として中心的である。
abstractwe propose LV-3DGS (Leafy Vegetables-3DGS), an optimized 3DGS-based framework tailored for the reconstruction of leafy vegetable scenes.
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.
Abstract To address the problem of fine branch identification and pruning decision for dormant apple trees, this study proposes a 3D point cloud branch recognition method integrating Neural Radiance Fields (NeRF) and the PointNeXt network. This method employs the neural radiance field theory to construct a point cloud model of apple trees, achieving fine detail representation and providing a high-precision, high-standard dataset for subsequent branch pruning experiments. First, a panoramic video is captured by circling the fruit tree, and a multi-view image sequence is obtained through frame sampling. Subsequently, the Structure from Motion (SfM) algorithm is employed for sparse reconstruction to recover the pose information of the images. On this basis, a neural radiance field model is trained. Hierarchical sampling is performed using ray casting, and the sampled points, combined with positional encoding, are fed into a multi-layer perceptron (MLP). The radiance field is then generated via volume rendering, from which a high-fidelity 3D point cloud model of the fruit tree is derived. Finally, the point cloud is processed using the PointNeXt semantic segmentation network to achieve the identification and segmentation of branches to be pruned and branches to be retained. To verify the effectiveness of the method, this study reconstructed point cloud models of dormant apple trees and selected 10 of them for experimental analysis. The algorithm achieved an average overall recognition accuracy of 75.15% and an average false negative rate (FNR) of 24.85%. The experimental results demonstrate that the proposed method constructs a 3D point cloud model with multi-scale, multi-modal, and high-precision phenotypic information at a relatively low cost. It not only overcomes the limitations of traditional 3D reconstruction methods, such as insufficient point cloud accuracy and difficulty in accurately identifying thin branches, but also effectively mitigates the high misrecognition rate observed in conventional branch recognition approaches. This provides technical support for unmanned agricultural machinery pruning in orchards and holds significant implications for achieving precision agriculture and sustainable development.
Why it matches plant phenotyping methodsNeRFとPointNeXtを用いてリンゴ樹の3D点群を構築し、剪定対象枝を認識・分割する手法が研究の中心であり、植物の形態・構造状態を直接推定して性能評価している。
abstractthis study proposes a 3D point cloud branch recognition method integrating Neural Radiance Fields (NeRF) and the PointNeXt network.
Land management and stewardship teams continue to lack the tools to capture 3D spatiotemporal insights of the ecosystems they oversee. For wildfire management at the wildland-urban interface, teams face challenges in capturing vegetation growth over time after a fuel reduction program and connecting seasonal changes to the vegetation distribution across the treated area. Current approaches rely on triangle meshes or point clouds generated from photogrammetry or LiDAR surveys on drones or hiked traverses. However, the difficulties in optimizing these meshes lead to large triangles that inadequately approximate the bulk vegetation shape, and the point cloud data is often too sparse for local plant-scale understanding. To address this gap, we extend recent machine learning-based computer graphics techniques in 3D Gaussian Splatting (3DGS) to reconstruct scenes at the wildland-urban interface from handheld imagery with sufficient detail to identify species, capture individual leaves, recover plant stature, and disambiguate overhanging plant individuals. We develop a new method to match the 3DGS reconstruction of these scenes across months, associating plant growth across seasons interactively in 3D. To achieve the centimeter-level matching, we adapt the Umeyama algorithm and the iterative closest point algorithm from point cloud maps to the 3DGS scene, leveraging the probabilistic interpretation of the 3D Gaussian data structure and robustly handling visual and geometric changes associated with vegetation phenology over time. We have applied our method to recent pile burns at Stanford’s Jasper Ridge ’Ootchamin ’Ooyakma Biological Preserve at monthly intervals. We demonstrate differences in ecological response where some piles featured the unexpected return of a rare and threatened bushmallow, and others remained more barren. This pile burn microcosm implicates the need for plant-level 3D spatiotemporal models to understand ecosystem recovery to fire mitigation practices. Please visit the project page for the spatiotemporal alignment video and more information: https://danineamati.github.io/burn-ecorecovery.github.io/ Please visit the Stanford Data Repository for the field data: https://doi.org/10.25740/ws901xs0162
Why it matches plant phenotyping methods植物の3D画像再構成と時系列位置合わせ手法を開発し、植物の個体・葉・草丈・成長を抽出することが中心であるため、植物フェノタイピング手法研究に該当します。
abstractwe extend recent machine learning-based computer graphics techniques in 3D Gaussian Splatting (3DGS) to reconstruct scenes at the wildland-urban interface from handheld imagery with sufficient detail to identify species, capture individual leaves, recover plant stature, and disambiguate overhanging plant individuals.
Abstract 3D Gaussian Splatting (3DGS) [KKLD23] has transformed novel‐view synthesis from RGB images, yet remains restricted to the visible spectrum. Many applications, including agricultural monitoring, rely on multi‐spectral imaging, where spectral camera alignment and scalability pose major challenges. We present MS‐Splatting—a multi‐spectral 3DGS framework enabling unified multi‐view consistent reconstruction and rendering across both visible and invisible spectra. Our key component is a neural color representation that encodes per‐primitive features shared across spectral bands, decoded through a shallow multi‐layer perceptron into spectrum‐specific radiance. By leveraging inter‐band correlations, this formulation enhances detail while reducing memory consumption compared to independent band modeling via per‐channel modeling with spherical harmonics. Our method enables accurate parallax‐free novel‐view vegetation index rendering for plant monitoring and enhances RGB novel view synthesis quality by exploiting details revealed through multi‐spectral bands. Our evaluation demonstrates that MS‐Splatting exceeds the current leading methods in both categories. In addition, we introduce a multi‐spectral dataset from aerial captures covering outdoor environments, specifically designed for evaluating these applications. We will release our code and dataset to facilitate further research. The project page is located at: https://meyerls.github.io/ms_splatting
Why it matches plant phenotyping methodsマルチスペクトル3D再構成法を開発し、植物モニタリング用の植生指数レンダリングを実現することが中心で、評価用データセットも提供している。
abstractWe present MS‐Splatting—a multi‐spectral 3DGS framework enabling unified multi‐view consistent reconstruction and rendering across both visible and invisible spectra.
Automated quantification of plant-level development from multi-plant greenhouse scenes requires separating individual plants from shared scene-level reconstructions and quantifying organ-level development, a challenge that single-plant acquisition workflows do not directly address. This study presents an end-to-end phenotyping pipeline built on 3D Gaussian Splatting (3DGS) and a post-reconstruction extraction framework, LCR-GS, designed to isolate plant instances from full greenhouse scenes without scene-specific model retraining. LCR-GS integrates zero-shot 2D cues with multi-view lifting, geometric clustering, and chromatic refinement to convert large scene-level reconstructions (~2M Gaussians) into compact per-plant subsets (~16K Gaussians). Experiments on greenhouse-grown muskmelon at the early vegetative stage demonstrate high plant-extraction precision (0.933) and strong organ-level instance segmentation (mean AP50 = 0.924). Plant height and leaf count are validated against manual measurements (height R² = 0.98, RMSE = 1.88 cm; leaf count R² = 0.86), whereas additional morphological traits, including leaf area, leaf area index, mean internode length, and stem node count, are reported as pipeline-derived descriptors for within-cohort comparison. By decoupling semantic inference from reconstruction, the pipeline reduces scene-scale data by over 99% and provides a practical route to derive compact per-plant 3D representations from multi-plant greenhouse imagery for downstream organ-level analysis.
Why it matches plant phenotyping methods3DGS画像から個体・器官を抽出し、植物形質を定量化するフェノタイピング手法の開発と検証が中心である。
abstractThis study presents an end-to-end phenotyping pipeline built on 3D Gaussian Splatting (3DGS) and a post-reconstruction extraction framework, LCR-GS, designed to isolate plant instances from full greenhouse scenes without scene-specific model retraining.
Reproduction assets foundThe paper's Data Availability Statement points to a public GitHub repository containing the muskmelon 3DGS phenotyping dataset (scenes, Gaussian-level plant/background annotations, and point-level organ labels) used in this study.Dataset · publicThe datasets presented in this study can be found in online repositories. The names of the repository/repositories and accession number(s) can be found below: https://github.com/bblabNTU/3dgs-muskmelon-phenotyping-dataset.Open asset ↗bblabNTU/3dgs-muskmelon-phenotyping-datasethtml-lines:485-547Plant phenotyping relevance match · UnverifiedOpenAlex · Crossref · Europe PMC · checked 5 Sept 2026
Introduction High-throughput field phenotyping (HTFP) holds great potential for elucidating the relationship between genomes and phenotypes. However, obtaining high-quality three-dimensional point cloud data of field populations and achieving single-plant phenotypic analysis remain challenging. Methods This study develops an integrated framework for field crop reconstruction based on 3D Gaussian splatting, incorporating a geometry-aware dynamic constraint algorithm to achieve instance segmentation and extract key phenotypic traits of individual plants. Using 3D Gaussian splatting technology, field-scale cotton population modeling is accomplished, generating dense 3D point clouds for regions of interest. Furthermore, the concept of a crop localization domain is proposed, establishing a longitudinal mapping that associates plant positional coordinates with long-term phenotypic attributes. Finally, through a dynamic spatial constraint mechanism, the accuracy and computational efficiency of instance segmentation for crop population point clouds are significantly improved, enabling rapid extraction of individual plant traits such as cotyledon node height, plant height, and leaf area. Results The results demonstrate that PhenotypeAI successfully reconstructed nine cotton populations with PSNR exceeding 30.0 dB. It successfully extracted regions of interest from 403 cotton plants, achieving an average F-score of 91.32% for instance segmentation and an average accuracy of 91.35%. The extracted traits—cotyledon node height, plant height, and leaf area—exhibited strong correlations with manual measurements, with coefficients of determination ( R 2 ) of 0.90, 0.91, and 0.91, respectively. Discussion The proposed method provides a low-cost solution for high-throughput field phenotypic analysis of field cotton and improves the efficiency of cotton breeding.
Why it matches plant phenotyping methods3D再構成と動的空間制約による個体セグメンテーションおよび形質抽出が研究の中心で、綿花の草丈・葉面積などを検証しているため。
abstractThis study develops an integrated framework for field crop reconstruction based on 3D Gaussian splatting, incorporating a geometry-aware dynamic constraint algorithm to achieve instance segmentation and extract key phenotypic traits of individual plants.
Abstract Maize（Zea mays L.） is an important crop, and improving its productivity is required even under challenging conditions such as labor shortages and uncertain climate fluctuations. One approach to enhancing yield is utilizing crop data for cultivation management and yield prediction. However, efficient acquisition of such data remains constrained by various limitations. In this study, we developed a non-contact and labor-efficient method for crop data acquisition by generating 3D models of maize at the ripening stage using Neural Radiance Fields (NeRF). Segmentation was performed on the obtained point clouds to estimate plant height, leaf area, and leaf angle. The coefficients of determination (R2) were 0.903, 0.954, and -0.521, respectively, demonstrating high accuracy for plant height and leaf area even at the ripening stage, while reducing the time required for data acquisition by 93% compared to manual measurements. Nevertheless, some manual operations−such as removing kernels and separating overlapping leaves−were still necessary, and full automation was not achieved. The main sources of error were identified as reconstruction errors in the base during scale adjustment, excessive removal of leaf sheaths, and the curvature of individual plants. Furthermore, we examined how measurement accuracy was influenced by factors such as the time of day and cultivar. The proposed method is expected to contribute to the practical implementation of a labor -saving 3D measurement technique that supports yield prediction and growth diagnosis in maize.
Why it matches plant phenotyping methodsNeRFによる3D再構成と点群セグメンテーションを開発し、トウモロコシの草丈・葉面積・葉角度を推定して精度と誤差要因を検証しており、表現型取得法が研究の中心である。
abstractwe developed a non-contact and labor-efficient method for crop data acquisition by generating 3D models of maize at the ripening stage using Neural Radiance Fields (NeRF).
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.
3D instance segmentation for laser scanning (LiDAR) point clouds remains a challenge in many remote sensing-related domains. Successful solutions typically rely on supervised deep learning and manual annotations, and consequently focus on objects that can be well delineated through visual inspection and manual labeling of point clouds. However, for tasks with more complex and cluttered scenes, such as in-field plant phenotyping in agriculture, such approaches are often infeasible. In this study, we tackle the task of in-field wheat head instance segmentation directly from terrestrial laser scanning (TLS) point clouds. To address the problem and circumvent the need for manual annotations, we propose a novel two-stage pipeline. To obtain the initial 3D instance proposals, the first stage uses 3D-to-2D multi-view projections, the Grounded SAM pipeline for zero-shot 2D object-centric segmentation, and multi-view label fusion. The second stage uses these initial proposals as noisy pseudo-labels to train a supervised 3D panoptic-style segmentation neural network. Our results demonstrate the feasibility of the proposed approach and show performance improvementsrelative to Wheat3DGS, a recent alternative solution for in-field wheat head instance segmentation without manual 3D annotations based on multi-view RGB images and 3D Gaussian Splatting, showcasing TLS as a competitive sensing alternative. Moreover, the results show that both stages of the proposed pipeline can deliver usable 3D instance segmentation without manual annotations, indicating promising, low-effort transferability to other comparable TLS-based point cloud segmentation tasks.
Why it matches plant phenotyping methodsTLS点群と深層学習によるコムギ穂の3D個体分割パイプラインを開発・評価しており、植物表現型取得手法が中心である。
abstractsuch as in-field plant phenotyping in agriculture
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 · UnverifiedOpenAlex · checked 13 Sept 2026
Aerial remote sensing efficiently surveys large areas, but accurate direct object-level measurement remains difficult in complex natural scenes. Advancements in 3D computer vision, particularly radiance field representations such as NeRF and 3D Gaussian splatting, can improve reconstruction fidelity from posed imagery. Nevertheless, direct aerial measurement of important attributes like tree diameter at breast height (DBH) remains challenging. Trunks in aerial forest scans are distant and sparsely observed in image views; at typical operating altitudes, stems may span only a few pixels. With these constraints, conventional reconstruction methods have inaccurate breast-height trunk geometry. TreeDGS is an aerial image reconstruction method that uses 3D Gaussian splatting as a continuous scene representation for trunk measurement. After SfM–MVS initialization and Gaussian optimization, we extract a dense point set from the Gaussian field using RaDe-GS’s depth-aware cumulative-opacity integration and associate each sample with a multi-view opacity reliability score. Then, we isolate trunk points and estimate DBH using opacity-weighted solid-circle fitting. Evaluated on 10 plots with field-measured DBH, TreeDGS reaches 4.79 cm RMSE (about 2.6 pixels at this GSD) and outperforms a LiDAR baseline (7.66 cm RMSE). This shows that TreeDGS can enable accurate, low-cost aerial DBH measurement.
Why it matches plant phenotyping methods樹木の胸高直径（DBH）という植物形態形質を、航空画像から3D再構成と信頼度重み付き推定で測定する手法を開発・検証しており、フェノタイピング手法が中心である。
abstractTreeDGS is an aerial image reconstruction method that uses 3D Gaussian splatting as a continuous scene representation for trunk measurement.
TeaField / plotNeRF / 3D Gaussian SplattingLiDAR / point cloudLeafWhole plant / canopy / plot / fieldCountingObject detection2D/3D reconstructionSegmentation
Accurate perception of tea buds is a fundamental prerequisite for intelligent and precise tea harvesting planning. However, in real tea plantation environments, reliable harvesting-oriented perception at the planning level remains highly challenging due to the small size of tea buds, severe occlusion, complex background clutter, and the lack of accurate three-dimensional spatial information. To address these challenges, we propose TeaNeRF, an integrated three-dimensional visual perception pipeline designed for harvesting-oriented tea bud analysis. Instead of treating detection, segmentation, and spatial analysis as independent tasks, TeaNeRF integrates sequential two-dimensional recognition, monocular depth estimation, and neural radiance field reconstruction into a coherent perception pipeline, allowing accurate spatial understanding of tea buds in complex natural scenes. It should be noted that the proposed integration is conducted at the perception-output level, where multiple modular components are connected through fixed interfaces, rather than through joint optimization or an end-to-end trainable formulation. The proposed framework combines an enhanced YOLO-based detector, prompt-guided segmentation, and monocular depth priors to guide NeRF-based three-dimensional reconstruction. By incorporating depth supervision and semantic-aware neural fields, TeaNeRF generates dense and geometrically consistent point clouds with reliable semantic separation. Quantitative evaluations show consistent improvements in reconstruction fidelity, as reflected by increased PSNR and reduced LPIPS across multiple tea tree scenes. Based on the reconstructed semantic point cloud, a three-dimensional clustering and geometric fitting strategy is further developed to enable tea bud counting and harvesting-oriented candidate point estimation at the perception level. Experiments conducted on a real-world dataset of 4,700 tea plantation images demonstrate that TeaNeRF improves detection accuracy (mAP@50 = 91.7%), segmentation quality (IoU = 0.640), and overall three-dimensional perception performance. Case-level counting results on representative tea trees indicate that the proposed 3D semantic point cloud-based approach can provide feasible tea bud counting behavior and consistent spatial guidance cues for downstream harvesting planning. By providing structured three-dimensional spatial information, including tea bud locations, counts, and harvesting-oriented candidate points, TeaNeRF offers practical perception-level outputs for downstream planning in automated tea harvesting systems.
Why it matches plant phenotyping methods茶芽の検出・セグメンテーション・3D再構成を統合し、茶芽の計数と3D位置推定を行う知覚パイプラインが研究の中心であり、単なる収穫対象の局在化を超えた器官形質の抽出を含む。
abstractwe propose TeaNeRF, an integrated three-dimensional visual perception pipeline designed for harvesting-oriented tea bud analysis.
NeRF / 3D Gaussian SplattingLiDAR / point cloudRGB / grayscaleWhole plant / canopy / plot / fieldOrgan identification2D/3D reconstructionSegmentation
Organ segmentation of plant point clouds is a prerequisite for the high-resolution and accurate extraction of organ-level phenotypic traits. Although the fast development of deep learning has boosted much research on segmentation of plant point clouds, the existing techniques for organ segmentation still face limitations in resolution, segmentation accuracy, and generalizability across various plant species. In this study, we proposed a novel approach called plant segmentation neural radiance fields (PlantSegNeRF), aiming to directly generate high-precision instance point clouds from multi-view RGB image sequences for a wide range of plant species. PlantSegNeRF performed two-dimensional (2D) instance segmentation on the multi-view images to generate instance masks for each organ with a corresponding instance identification (ID). The multi-view instance IDs corresponding to the same plant organ were then matched and refined using a specially designed instance matching (IM) module. The instance NeRF was developed to render an implicit scene containing color, density, semantic and instance information, which was ultimately converted into high-precision plant instance point clouds based on volume density. The results proved that in semantic segmentation of point clouds, PlantSegNeRF outperformed the commonly used methods, demonstrating an average improvement of 16.1 %, 18.3 %, 17.8 %, and 24.2 % in precision, recall, F1-score, and intersection over union (IoU) compared to the second-best results on structurally complex datasets. More importantly, PlantSegNeRF exhibited significant advantages in instance segmentation. Across all plant datasets, it achieved average improvements of 11.7 %, 38.2 %, 32.2 % and 25.3 % in mean precision (mPrec), mean recall (mRec), mean coverage (mCov), and mean weighted coverage (mWCov), respectively. Furthermore, PlantSegNeRF demonstrates superior few-shot, cross-species performance, requiring only multi-view images of few plants to train models applicable to specific or similar varieties. This study extends organ-level plant phenotyping and provides a high-throughput way to supply high-quality 3D data for developing large-scale artificial intelligence (AI) models in plant science. • A comprehensive dataset of well-labeled two-dimensional (2D) images and point clouds dataset of plants was established, including various varieties and growth stages. 50 plant samples were collected for each type. • A novel multi-view image instance matching (IM) module was proposed to align plant organ instance identifications (IDs) across different viewpoints, serving as the foundation for organ-level instance segmentation. • A multi-channel instance neural radiance fields (NeRF) module with encoding color, semantic, and instance information was developed to achieve high-precision mapping of 2D image colors, semantics, and aligned instances into 3D space, enabling point cloud background removal and fine-grained segmentation of plant organs.
Why it matches plant phenotyping methods植物器官の3D点群再構成・インスタンス分割を開発し、セグメンテーション性能を検証する手法研究であり、器官レベル表現型抽出を直接支援するため。
abstractwe proposed a novel approach called plant segmentation neural radiance fields (PlantSegNeRF), aiming to directly generate high-precision instance point clouds from multi-view RGB image sequences for a wide range of plant species.
Field / plotLaboratory / benchtopNeRF / 3D Gaussian SplattingPhotogrammetry / SfM / MVSLiDAR / point cloudRGB-D / ToFWhole plant / canopy / plot / field2D/3D reconstruction
Three-dimensional point cloud (3DPC) data capture detailed geometric and structural plant traits beyond the capability of 2D imaging. When combined with artificial intelligence (AI), it offers a powerful, non-invasive tool for plant phenotyping, which is crucial for driving advancements in plant breeding and agriculture. However, challenges related to data complexity, limited datasets, and model generalization hinder 3DPC’s widespread adoption. To provide a comprehensive overview and guide future research in this area, we conducted a systematic literature review (SLR) following Preferred Reporting Items for Systematic reviews and Meta-Analyses (PRISMA) guidelines by analysing 381 papers published between January 2017 and October 2025 from major databases. Our review examines the advantages, current status, limitations, and future directions of AI applications in 3DPC-based plant phenotyping. Our findings indicate a rapid increase in publications since 2022, with deep learning (DL) methods, especially pointwise MLP-based networks, driving much of this growth, with a notable recent surge in Transformer-based, Graph-based, and particularly Hybrid models that combine their strengths. Furthermore, novel methods like Neural Radiance Fields (NeRF) and 3D Gaussian Splatting (3DGS) are emerging as powerful tools for 3D reconstruction and scene synthesis. Time-of-Flight (ToF) and Structure from Motion and Multi-View Stereo (SfM-MVS) technologies remain the predominant 3DPC data acquisition techniques. Research in this area focuses on trees/shrubs and cereals, typically involving single-species studies. Although the overall use of public datasets remains low (18.9%), their adoption has significantly increased since 2020. Key limitations identified include: (1) a lack of standardized data collection and formats, (2) insufficient model robustness and generalization, especially from lab to field, (3) high computational demands, and (4) a reliance on species-specific models. The future of AI-driven 3DPC phenotyping hinges on overcoming these bottlenecks. Priority should be given to: developing field-deployable, computationally efficient models; exploring the potential of the foundation model; establishing diverse and standardized public datasets; and strengthening the integration of 3D phenomics with genomics to bridge the genotype-to-phenotype gap. This review provides a foundational roadmap to guide research in plant phenomics, crop breeding, and plant science.
Why it matches plant phenotyping methods3D点群とAIによる植物形質取得・解析を中心に扱う体系的レビューであり、植物フェノタイピング手法のレビューとして明確に適格。
titleAI-driven 3D point cloud analysis in plant phenotyping: A Systematic Review
Field / plotNeRF / 3D Gaussian SplattingPhotogrammetry / SfM / MVSLiDAR / point cloudLeafStem / branchWhole plant / canopy / plot / fieldMorphology / geometry measurement2D/3D reconstructionSkeletonization / topology
Saplings are key indicators of forest regeneration and overall forest health. However, their fine-scale architectural traits are difficult to capture with existing 3D sensing methods, which make quantitative evaluation difficult. Terrestrial Laser Scanners (TLS), Mobile Laser Scanners (MLS), or traditional photogrammetry approaches poorly reconstruct thin branches, dense foliage, and lack the scale consistency needed for long-term monitoring. Implicit 3D reconstruction methods such as Neural Radiance Fields (NeRF) and 3D Gaussian Splatting (3DGS) are promising alternatives, but cannot recover the true scale of a scene and lack any means to be accurately geo-localised. In this paper, we present a pipeline which fuses NeRF, LiDAR SLAM, and GNSS to enable repeatable, geo-localised ecological monitoring of saplings. Our system proposes a three-level representation: (i) coarse Earth-frame localisation using GNSS, (ii) LiDAR-based SLAM for centimetre-accurate localisation and reconstruction, and (iii) NeRF-derived object-centric dense reconstruction of individual saplings. This approach enables repeatable quantitative evaluation and long-term monitoring of sapling traits. Our experiments in forest plots in Wytham Woods (Oxford, UK) and Evo (Finland) show that stem height, branching patterns, and leaf-to-wood ratios can be captured with increased accuracy as compared to TLS. We demonstrate that accurate stem skeletons and leaf distributions can be measured for saplings with heights between 0.5m and 2m in situ, giving ecologists access to richer structural and quantitative data for analysing forest dynamics.
Why it matches plant phenotyping methodsNeRF・LiDAR SLAM・GNSSを融合した幼木の3D再構成・定位パイプラインを開発し、樹高、分枝、葉対木質比などの植物形質をTLSと比較検証しており、表現型取得手法が中心である。
abstractIn this paper, we present a pipeline which fuses NeRF, LiDAR SLAM, and GNSS to enable repeatable, geo-localised ecological monitoring of saplings.
• 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.
Modeling the time-varying 3D appearance of plants during growth poses unique challenges: unlike most dynamic scenes, plants continuously generate new geometry as they expand, branch, and differentiate. Existing dynamic scene representations are ill-suited to this setting: deformation fields provide insufficient constraints to yield physically plausible scene dynamics, and 4D Gaussian splatting represents the same physical structures with different Gaussian primitives at different times, breaking temporal consistency. We introduce GrowFlow, a dynamic representation that couples 3D Gaussian primitives with a neural ordinary differential equation to model plant growth as a continuous flow field over geometric parameters (position, scale, and orientation). Our representation enables consistent appearance rendering and models nonlinear, continuous-time growth dynamics with full temporal correspondences for every primitive. To initialize a sufficient set of Gaussian primitives, we first reconstruct the mature plant and then learn a reverse-growth process, effectively simulating the plant's developmental history in reverse. GrowFlow achieves superior image quality and geometric coherence compared to prior methods on a new, multi-view timelapse dataset of plant growth, and provides the first temporally coherent representation for appearance modeling of growing 3D structures.
Why it matches plant phenotyping methods植物の成長を対象に、4D再構成と連続的な成長表現を開発し、幾何学的整合性と画像品質を既存手法・新規データセットで比較評価しているため、植物フェノタイピング手法が中心です。
abstractWe introduce GrowFlow, a dynamic representation that couples 3D Gaussian primitives with a neural ordinary differential equation to model plant growth as a continuous flow field over geometric parameters (position, scale, and orientation).
Analyzing three-dimensional (3D) phenotypic parameters of maize seedlings is of significant importance for maize cultivation and selection. However, existing methods often struggle to balance cost, efficiency, and accuracy, particularly when capturing the complex morphology of seedlings characterized by slender stems. To address these issues, this study proposes a novel end-to-end automated framework for extracting phenotypes using only consumer-grade RGB cameras. The pipeline initiates with Instant-NGP to rapidly reconstruct dense point clouds, establishing the 3D data foundation for phenotypic extraction. Subsequently, we formulate a directed topological graph-based mechanism. By mathematically defining bifurcation constraints via vector analysis, this mechanism guides a depth-first traversal strategy to explicitly disentangle stem and leaf skeletons. Building upon these decoupled skeletons, organ-level point cloud segmentation is achieved through constraint-based expansion, followed by density-based spatial clustering (DBSCAN) to detect individual leaves. Algorithms combining point cloud geometry with 3D Euclidean distance are also implemented to calculate key phenotypes including plant height and stem width. Finally, single-leaf skeleton fitting is used to estimate leaf length, and principal component analysis (PCA) is adopted to determine the stem–leaf angle, realizing the comprehensive automatic extraction of maize seedling phenotypes. Experiments show that the proposed method achieves high accuracy in extracting key phenotypic parameters. The mean relative errors for plant height, stem width, leaf length, stem-leaf angle, and leaf area are 0.76%, 2.93%, 1.26%, 2.13%, and 3.33%, respectively. Compared with existing methods as far as we know, the proposed method significantly improves extraction efficiency by reducing the processing time per plant to within 5 min while maintaining such high accuracy.
Why it matches plant phenotyping methodsRGBカメラとNeRF・点群処理・骨格解析を統合し、トウモロコシ幼苗の形質を自動抽出する手法を開発・精度評価した研究であり、フェノタイピング手法が中心である。
abstractthis study proposes a novel end-to-end automated framework for extracting phenotypes using only consumer-grade RGB cameras.
Fine-grained 3D phenotypic analysis of rice plays a vital role in rice breeding and yield estimation. However, a comprehensive rice data acquisition and segmentation pipeline is still lacking. While Neural Radiance Fields (NeRF) have shown impressive results in crop-level 3D reconstruction, their high sensitivity to data volume and camera viewpoints often leads to reconstruction failures for rice. In addition, the large-scale rice point clouds, coupled with heavy occlusion and visual similarity among grains, pose significant challenges for fine-grained trait extraction. To address the challenge of reconstructing rice point clouds under low-quality data conditions, we propose a novel method named Multi-Scale NeRF(MSNeRF). This method incorporates a structure-detail collaborative reconstruction mechanism and a dynamic initialization density scheduling strategy. Furthermore, we introduce a multimodal and multitask rice dataset (MMR) as a benchmark resource for future research. For rice point cloud segmentation, we develop Vision Rice Knowledge Graph Network(VRKGNet), which comprises an image segmentation module, a projection module, and a point cloud segmentation module enhanced with a Transformer to enlarge the receptive field. VRKGNet performs standalone point cloud segmentation and integrates image segmentation results from multiple viewpoints as prior knowledge to enhance semantic and instance-level segmentation. Extensive experiments demonstrate that MSNeRF achieves high-fidelity point cloud reconstruction with as few as 10 viewpoints. VRKGNet achieves superior rice plant segmentation with a semantic segmentation mIoU of 88.79% and an instance segmentation AP 25 of 84.55%, outperforming mainstream algorithms.
Why it matches plant phenotyping methods米の3D形質取得・再構成・分割を中核とする手法開発であり、データセット／ベンチマークも提供しているため、植物フェノタイピング手法文献に該当する。
abstractwe propose a novel method named Multi-Scale NeRF(MSNeRF)
Reproduction assets foundThe paper's authors explicitly state that the source code for MSNeRF and VRKGNet is publicly available on GitHub with testing scripts and test cases to reproduce the main results. The MMR dataset itself is only available upon request from the corresponding author, so it does not qualify as a public asset.Code · publicof Hefei Artificial Intelligence Breeding Accelerator Co. Ltd. ( NB2024005-02 ).
Data availability
The source code for the proposed methods, MSNeRF and VRKGNet, is publicly available on GitHub. The released repositories contain testing scripts and test cases used to reproduce the main results presented in this paper:
•
MSNeRF : https://github.com/qfwysw/MSNeRF.git
•
VRKGNet : https://github.com/qfwysw/VRKGNet.git
The datasets used in the experiments are available from the corresponding author upon reasonable request. For access or further inquiries, please contact the corresponding author.
Declaration of competing interest
The authors declare that they have no known competing financial iOpen asset ↗https://github.com/qfwysw/MSNeRF.gitlines:620-663Code · publicator Co. Ltd. ( NB2024005-02 ).
Data availability
The source code for the proposed methods, MSNeRF and VRKGNet, is publicly available on GitHub. The released repositories contain testing scripts and test cases used to reproduce the main results presented in this paper:
•
MSNeRF : https://github.com/qfwysw/MSNeRF.git
•
VRKGNet : https://github.com/qfwysw/VRKGNet.git
The datasets used in the experiments are available from the corresponding author upon reasonable request. For access or further inquiries, please contact the corresponding author.
Declaration of competing interest
The authors declare that they have no known competing financial interests or personal relationships that could haveOpen asset ↗https://github.com/qfwysw/VRKGNet.gitlines:620-663Plant phenotyping relevance match · UnverifiedCrossref · OpenAlex · checked 15 Sept 2026
Sugarcane is an important economic crop, and key phenotypic traits such as plant height and leaf area play a crucial role in yield potential assessment and breeding selection. However, the quantification of these traits currently relies mainly on inefficient and destructive manual measurements, making it difficult to achieve continuous monitoring of plant growth. To address this limitation, this study integrates a YOLOv8x-seg instance segmentation model with 3D Gaussian Splatting (3DGS) and proposes a non-contact, high-precision 3D phenotyping method based on low-cost data acquisition using a smartphone. Multi-view RGB images are first processed using YOLOv8x-seg to extract plant foreground masks, which are then used as inputs for 3DGS-based reconstruction to generate 3D models. Plant height is automatically measured from the reconstructed models, while leaf area extraction involves a semi-automatic workflow combining image processing and manual steps. Experimental results demonstrate that the proposed approach enables accurate trait estimation, achieving a coefficient of determination (R2) of 0.9644 for plant height estimation (evaluated on a subset of 15 plants, with a mean absolute percentage error of approximately 1.5%) and an R2 of 0.8551 for leaf area estimation (validated on 10 plants). Ground-truth plant height was measured using a telescopic measuring rod, and leaf area was determined through destructive measurement with a leaf area meter (LI-COR Model LI-3000A). Ground-truth plant height values were obtained using a telescopic measuring rod, and leaf area was determined through destructive measurement with a leaf area meter (LI-COR Model LI-3000A). This method demonstrates the feasibility of using consumer-grade devices for high-fidelity 3D phenotyping and offers an effective approach for high-throughput sugarcane breeding applications.
Why it matches plant phenotyping methodsスマートフォン画像、インスタンスセグメンテーション、3D再構成を統合し、サトウキビの草高・葉面積を自動／半自動推定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。
abstractproposes a non-contact, high-precision 3D phenotyping method based on low-cost data acquisition using a smartphone
Field / plotNeRF / 3D Gaussian SplattingPhotogrammetry / SfM / MVSLiDAR / point cloudLeafWhole plant / canopy / plot / field2D/3D reconstructionImage / point-cloud registrationGrowth / time-series analysisTracking
Land management and stewardship teams continue to lack the tools to capture 3D spatiotemporal insights of the ecosystems they oversee. For wildfire management at the wildland-urban interface, teams face challenges in capturing vegetation growth over time after a fuel reduction program and connecting seasonal changes to the vegetation distribution across the treated area. Current approaches rely on triangle meshes or point clouds generated from photogrammetry or LiDAR surveys on drones or hiked traverses. However, the difficulties in optimizing these meshes lead to large triangles that inadequately approximate the bulk vegetation shape, and the point cloud data is often too sparse for local plant-scale understanding. To address this gap, we extend recent machine learning-based computer graphics techniques in 3D Gaussian Splatting (3DGS) to reconstruct scenes at the wildland-urban interface from handheld imagery with sufficient detail to identify species, capture individual leaves, recover plant stature, and disambiguate overhanging plant individuals. We develop a new method to match the 3DGS reconstruction of these scenes across months, associating plant growth across seasons interactively in 3D. To achieve the centimeter-level matching, we adapt the Umeyama algorithm and the iterative closest point algorithm from point cloud maps to the 3DGS scene, leveraging the probabilistic interpretation of the 3D Gaussian data structure and robustly handling visual and geometric changes associated with vegetation phenology over time. We have applied our method to recent pile burns at Stanford’s Jasper Ridge ’Ootchamin ’Ooyakma Biological Preserve at monthly intervals. We demonstrate differences in ecological response where some piles featured the unexpected return of a rare and threatened bushmallow, and others remained more barren. This pile burn microcosm implicates the need for plant-level 3D spatiotemporal models to understand ecosystem recovery to fire mitigation practices. Please visit the project page for the spatiotemporal alignment video and more information: https://danineamati.github.io/burn-ecorecovery.github.io/
Why it matches plant phenotyping methods植物の3D画像再構成と時系列位置合わせを開発し、植物の個体・葉・草丈・成長を抽出することが中心であるため、植物フェノタイピング手法として含める。
abstractwe extend recent machine learning-based computer graphics techniques in 3D Gaussian Splatting (3DGS) to reconstruct scenes at the wildland-urban interface from handheld imagery with sufficient detail to identify species, capture individual leaves, recover plant stature, and disambiguate overhanging plant individuals.
Strawberries are among the most economically significant fruits in the United States, generating over $2 billion in annual farm-gate sales and accounting for approximately 13% of the total fruit production value. Plant phenotyping plays a vital role in selecting superior cultivars by characterizing plant traits such as morphology, canopy structure, and growth dynamics. However, traditional plant phenotyping methods are time-consuming, labor-intensive, and often destructive. Recently, neural rendering techniques, notably Neural Radiance Fields (NeRF) and 3D Gaussian Splatting (3DGS), have emerged as powerful frameworks for high-fidelity 3D reconstruction. By capturing a sequence of multi-view images or videos around a target plant, these methods enable non-destructive reconstruction of complex plant architectures. Despite their promise, most current applications of 3DGS in agricultural domains reconstruct the entire scene, including background elements, which introduces noise, increases computational costs, and complicates downstream trait analysis. To address this limitation, we propose a novel object-centric 3D reconstruction framework incorporating a preprocessing pipeline that leverages the Segment Anything Model v2 (SAM-2) and alpha channel background masking to achieve clean strawberry plant reconstructions. This approach produces more accurate geometric representations while substantially reducing computational time. With a background-free reconstruction, our algorithm can automatically estimate important plant traits, such as plant height and canopy width, using DBSCAN clustering and Principal Component Analysis (PCA). Experimental results show that our method outperforms conventional pipelines in both accuracy and efficiency, offering a scalable and non-destructive solution for strawberry plant phenotyping.
Why it matches plant phenotyping methods植物の3D再構成、背景除去、クラスタリングを統合し、草丈やキャノピー幅を自動推定するフェノタイピング手法の開発が中心である。
abstractwe propose a novel object-centric 3D reconstruction framework incorporating a preprocessing pipeline that leverages the Segment Anything Model v2 (SAM-2) and alpha channel background masking to achieve clean strawberry plant reconstructions.
Aerial remote sensing efficiently surveys large areas, but accurate direct object-level measurement remains difficult in complex natural scenes. Advancements in 3D computer vision, particularly radiance field representations such as NeRF and 3D Gaussian splatting, can improve reconstruction fidelity from posed imagery. Nevertheless, direct aerial measurement of important attributes like tree diameter at breast height (DBH) remains challenging. Trunks in aerial forest scans are distant and sparsely observed in image views; at typical operating altitudes, stems may span only a few pixels. With these constraints, conventional reconstruction methods have inaccurate breast-height trunk geometry. TreeDGS is an aerial image reconstruction method that uses 3D Gaussian splatting as a continuous scene representation for trunk measurement. After SfM--MVS initialization and Gaussian optimization, we extract a dense point set from the Gaussian field using RaDe-GS's depth-aware cumulative-opacity integration and associate each sample with a multi-view opacity reliability score. Then, we isolate trunk points and estimate DBH using opacity-weighted solid-circle fitting. Evaluated on 10 plots with field-measured DBH, TreeDGS reaches 4.79 cm RMSE (about 2.6 pixels at this GSD) and outperforms a LiDAR baseline (7.66 cm RMSE). This shows that TreeDGS can enable accurate, low-cost aerial DBH measurement .
Why it matches plant phenotyping methods樹木のDBHという明示的な植物形質を、航空画像と3D Gaussian splattingから推定する新規手法を開発し、実測値およびLiDARと比較検証しているため、方法中心の植物フェノタイピング研究として含める。
abstractTreeDGS is an aerial image reconstruction method that uses 3D Gaussian splatting as a continuous scene representation for trunk measurement.
Achieving an optimal plant architecture is a key objective in cotton breeding for enhancing yield potential, and accurate quantification of the fruit branch angle (FBA) is essential for understanding genotype–phenotype relationships and advancing ideotype breeding. However, in-field FBA measurement remains technically challenging due to severe occlusion, variable illumination, and background interference. To overcome these limitations, we propose a streamlined 3D phenotyping framework that integrates 3D Gaussian Splatting (3DGS) with a novel structural segmentation model, the Linear Point Cloud Reverse Model (LPCRM). The framework decomposes reconstructed cotton point clouds into linear micro-elements using RANSAC, followed by geometric clustering via K-means to identify and separate the main stem and fruit branches. This process operates without topological priors or large annotated datasets. Model fidelity assessment shows that 80% of point pairs between the LPCRM and the original 3DGS reconstruction exhibit Euclidean distances ≤ 0.5 cm. Phenotypic validation using 268 fruit branches from 25 cultivars demonstrates high measurement accuracy, achieving an R² of 0.874 and an RMSE of 4.01° for FBA extraction. Plant height estimation also shows strong agreement with manual measurements (R² = 0.915, RMSE = 3.858). Overall, this study presents a lightweight and robust solution for extracting 3D structural traits of field-grown cotton. The proposed framework reduces data dependency, adapts well to complex field conditions, and offers an efficient approach for high-throughput phenotyping and cotton ideotype breeding.
Why it matches plant phenotyping methods綿花の果枝角度などの3D植物形質を抽出する画像ベース表現型解析フレームワークを開発し、複数品種・枝で精度検証しており、方法開発と技術検証が研究の中心である。
abstractwe propose a streamlined 3D phenotyping framework that integrates 3D Gaussian Splatting (3DGS) with a novel structural segmentation model, the Linear Point Cloud Reverse Model (LPCRM).
Semantic reconstruction of agricultural scenes plays a vital role in tasks such as phenotyping and yield estimation. However, traditional approaches based on manual scanning or fixed camera setups remain a major bottleneck, while active-mapping methods based solely on occupancy grids are too coarse for accurate trait estimation. To address this gap, we propose an active 3D reconstruction framework for horticultural environments using a mobile manipulator. The system integrates OctoMap with 3D Gaussian Splatting to enable accurate and efficient target-aware mapping. A low-resolution OctoMap provides probabilistic occupancy information for informative viewpoint selection and collision-free planning, while 3D Gaussian Splatting leverages geometric, photometric, and semantic information to optimize 3D Gaussians for high-fidelity scene reconstruction. We further introduce a robust mapping strategy that mitigates semantic segmentation and depth noise, together with a background pruning method that reduces memory and computational cost. We validate our framework across simulated, laboratory, and real greenhouse scenes, showing consistent improvements across three state-of-the-art Gaussian Splatting backbones. In simulation, where ground-truth geometry is available, our approach outperforms occupancy-based mapping in both reconstruction accuracy and runtime efficiency: compared with a 0.01m-resolution OctoMap, it doubles the fruit-level F1 score under noisy conditions while achieving up to a threefold reduction in runtime. Beyond simulation, novel-view synthesis quality also improves consistently in laboratory and real greenhouse environments, with PSNR and mIoU improving by up to 1.5 dB and 18%, respectively. Finally, the reconstructed semantic maps enable fruit counting and volume estimation with accuracies approaching 80%.
Why it matches plant phenotyping methods園芸ロボット向けの3D再構成・能動マッピング手法を開発し、果実の計数・体積推定という植物形質の取得に適用・検証しているため、フェノタイピング手法が中心的です。
titleOctoSplat: Hybrid OctoMap-Gaussian Splatting for Active Semantic Mapping and Phenotyping with Horticultural Robots
Reproduction assets foundThe paper's supplementary material is hosted on the authors' public project page (jrcuaranv.github.io/octosplat), and the authors state that all code and data are publicly available. The SimSense repository is a third-party depth-sensor simulator tool, not a paper-specific asset.Code · publicAll code and data are publicly available to facilitate reproducibility.Open asset ↗lines:59-163Code / dataset availability confirmedEurope PMC · OpenAlex · checked 15 Sept 2026
Fruit growth has long been described using single- or double-sigmoid curves; however, these temporal models cannot fully capture the spatial heterogeneity that ultimately shapes a fruit. Here, we present a three-dimensional analysis pipeline that non-destructively tracks spatial fruit growth dynamics from field-collected imagery. Surface landmarks were drawn, and video recordings were taken throughout development for three pome fruits, apple ( Malus × domestica ), Japanese pear ( Pyrus pyrifolia ) and European pear ( Pyrus communis ), and two drupe fruits, peach ( Prunus persica ) and Japanese apricot ( Prunus mume ), to track their motion. Using 3D Gaussian splatting, we successfully reconstructed 3D models of the fruits, and the landmark displacement could be measured with high accuracy, with R 2 ≥ 0.98 when compared to manual recordings. We found a common spatial growth gradient in the longitudinal growth shared in the pomes and drupes of the Rosaceae; proximal (stem-end) regions exhibited more pronounced growth than the distal (stylar) end. An exception was found in European pear 'Bartlett,' which showed relatively vigorous growth in the distal region, explaining its distinct shape with expanded distal end. Transverse expansion varied far less than longitudinal expansion, with a possible association with initial fruit morphology. Inter-fruit growth variability peaked in the fastest-growing regions, particularly in the distal area of the European pear, highlighting the link between growth vigor and phenotypic variance. These results provide foundational insights into the developmental dynamics of both pome and drupe fruits of the Rosaceae family, contributing to the optimization of fruit size, shape, and uniformity.
Why it matches plant phenotyping methods3D画像解析パイプラインと3D Gaussian Splattingを用いて果実の空間的成長を非破壊計測し、手動記録との精度比較で検証しているため、植物表現型取得法が中心である。
abstractHere, we present a three-dimensional analysis pipeline that non-destructively tracks spatial fruit growth dynamics from field-collected imagery.
Reproduction assets foundThe authors deposited a subset of the 3DGS-reconstructed fruit models (the paper's phenotyping outputs) on Figshare with a public DOI; additional data only on request. No author analysis code or raw imagery deposit is stated.Dataset · publicFootnotes
Appendix A
Supplementary data to this article can be found online at https://doi.org/10.1016/j.plaphe.2026.100166 .
Appendix A.
Supplementary data
The following is the Supplementary data to this article:
Multimedia component 1
Multimedia component 1
Data availability
A subset of the generated 3D models is available at https://doi.org/10.6084/m9.figshare.30854579 , where the quality of the 3DGS reconstructions and the marking/measurement procedure can be examined. Additional data may be provided upon reasonable request to the corresponding author.
ReferencesOpen asset ↗figshare · 10.6084/m9.figshare.30854579lines:151-171Code / dataset availability confirmedarXiv · OpenAlex · checked 15 Sept 2026
AppleCottonPearField / plotNeRF / 3D Gaussian SplattingFruitCounting2D/3D reconstructionSegmentation
Rigorous crop counting is crucial for effective agricultural management and informed intervention strategies. However, in outdoor field environments, partial occlusions combined with inherent ambiguity in distinguishing clustered crops from individual viewpoints poses an immense challenge for image-based segmentation methods. To address these problems, we introduce a novel crop counting framework designed for exact enumeration via 3D instance segmentation. Our approach utilizes 2D images captured from multiple viewpoints and associates independent instance masks for neural radiance field (NeRF) view synthesis. We introduce crop visibility and mask consistency scores, which are incorporated alongside 3D information from a NeRF model. This results in an effective segmentation of crop instances in 3D and highly-accurate crop counts. Furthermore, our method eliminates the dependence on crop-specific parameter tuning. We validate our framework on three agricultural datasets consisting of cotton bolls, apples, and pears, and demonstrate consistent counting performance despite major variations in crop color, shape, and size. A comparative analysis against the state of the art highlights superior performance on crop counting tasks. Lastly, we contribute a cotton plant dataset to advance further research on this topic.
Why it matches plant phenotyping methodsNeRFと3Dインスタンスセグメンテーションを用いて作物個体・器官数を推定する画像ベース表現型計測手法を開発・検証しており、方法が研究の中心である。
abstractwe introduce a novel crop counting framework designed for exact enumeration via 3D instance segmentation.
Reproduction assets foundThe paper contributes a public infield cotton plant dataset (8 plants, ~150 iPhone images each, ground-truth boll counts, SAM instance masks) and states that source code, dataset, and multimedia are available at the authors' public project page, which is an allowed URL. The spectacularai GitHub URL is a generic third-pDataset · publicthat incorporates crop visibility
and mask consistency, enabling robustness against occlusions and annotation
discrepancies.
•
We release a public infield cotton plant dataset designed for 3D
rendering and cotton boll counting tasks.
The source code, dataset, and multimedia material associated with this project
can be found at
https://robotic-vision-lab.github.io/cropnerf .
II Related Work
II-A Image-Based Techniques
Image-based methods typically employ object detection to identify crops within
images. For example, Chen et al. [ 4 ] utilized multiple
convolutional neural networks (CNNs) to map input images to total fruit counts.
Similarly, Häni et al. [ 5 ] formulated crop counting as a
multOpen asset ↗lines:108-187Plant phenotyping relevance match · UnverifiedOpenAlex · checked 15 Sept 2026
SoybeanNeRF / 3D Gaussian SplattingLiDAR / point cloudSegmentationTracking
Abstract has not been obtained from indexed metadata or an accessible article page.
Why it matches plant phenotyping methods成熟ダイズ個体を対象とする点群インスタンスセグメンテーション・パイプラインの開発であり、植物画像から個体を分離・再構成する手法が中心です。
titleSoybeanInsGS: A high-precision, data-efficient point cloud instance segmentation pipeline for mature soybean plants via cross-view instance tracking and instance-aware 3DGS
High-precision plant phenotyping requires efficient 3D reconstruction with high fidelity, yet existing methods such as MVS and NeRF all have problems of feature dependence and error accumulation during 3D reconstruction, which leads to geometric distortion in reconstruction and restricts the reconstruction efficiency. To address this bottleneck, this study first determined the multi-view image acquisition strategy. Further, based on the self-built multi-view dataset of chili peppers, it proposed an algorithm for efficient and high-fidelity 3D reconstruction of complex plant structures through global adaptive pose optimization and gaussian splash rendering technology, referred to as the GAPose-GS algorithm. Experimental results indicate that the Peak Signal-to-Noise Ratio ( PSNR ) improves by 52.0 %, 26.4 %, and 4.2 % compared to NeRF, Instant-NGP, and 3D Gaussian Splatting respectively. Additionally, the Structural Similarity Index Measure ( SSIM ) increases by 22.9 %, 12.8 %, and 4.3 % respectively over above methods. The point cloud data reconstructed based on this algorithm also has advantages in the measurement of phenotypic parameters. Compared with the actual measured values, the R² of the phenotypic parameters such as pepper plant height, canopy width, and leafstalk angle obtained in this study are 0.997, 0.954 and 0.978 respectively, and the RMSE are 0.236 cm, 1.082 cm and 2.344° respectively, and the MAE are 0.209 cm, 0.880 cm and 1.965° respectively. The accuracy was significantly better than that of the existing phenotypic calculation methods. Verification across different growth stages of wheat and maize was performed universally, with all errors remaining below 1.1 %, providing new ideas and technologies for high-precision, low-cost, and high-throughput crop phenotypic research.
Why it matches plant phenotyping methods植物の多視点画像から3D再構成し、草丈・群落幅・葉柄角などの形質を推定する手法を開発・比較検証しており、フェノタイピング手法が中心的である。
abstractit proposed an algorithm for efficient and high-fidelity 3D reconstruction of complex plant structures through global adaptive pose optimization and gaussian splash rendering technology, referred to as the GAPose-GS algorithm.
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 · UnverifiedOpenAlex · arXiv · checked 6 Sept 2026
We present a method for jointly recovering the appearance and internal structure of botanical plants from multi-view images based on 3D Gaussian Splatting (3DGS). While 3DGS exhibits robust reconstruction of scene appearance for novel-view synthesis, it lacks structural representations underlying those appearances (e.g., branching patterns of plants), which limits its applicability to tasks such as plant phenotyping. To achieve both high-fidelity appearance and structural reconstruction, we introduce GaussianPlant, a hierarchical 3DGS representation, which disentangles structure and appearance. Specifically, we employ structure primitives (StPs) to explicitly represent branch and leaf geometry, and appearance primitives (ApPs) to the plants' appearance using 3D Gaussians. StPs represent a simplified structure of the plant, i.e., modeling branches as cylinders and leaves as disks. To accurately distinguish the branches and leaves, StP's attributes (i.e., branches or leaves) are optimized in a self-organized manner. ApPs are bound to each StP to represent the appearance of branches or leaves as in conventional 3DGS. StPs and ApPs are jointly optimized using a re-rendering loss on the input multi-view images, as well as the gradient flow from ApP to StP using the binding correspondence information. We conduct experiments to qualitatively evaluate the reconstruction accuracy of both appearance and structure, as well as real-world experiments to qualitatively validate the practical performance. Experiments show that the GaussianPlant achieves both high-fidelity appearance reconstruction via ApPs and accurate structural reconstruction via StPs, enabling the extraction of branch structure and leaf instances.
Why it matches plant phenotyping methods植物の多視点画像から枝構造と葉インスタンスを再構成・抽出する手法を開発しており、植物フェノタイピングへの適用と形態情報の抽出が中心的な技術貢献である。
abstractWe present a method for jointly recovering the appearance and internal structure of botanical plants from multi-view images based on 3D Gaussian Splatting (3DGS).
Accurate and efficient plant phenotyping is essential for modern precision agriculture. as it provides reliable information for seedling quality evaluation, early detection of plant stress, and data support for crop breeding and yield prediction. Traditional three-dimensional (3D) reconstruction and analysis methods are often costly and time-consuming, because they usually depend on expensive laser scanning devices or require many input images. Even with these resources, they often fail to capture fine plant structures such as leaves and branches, which limits their application in seedling monitoring. To address these challenges, we propose an integrated framework that combines neural radiance fields (NeRFs) for high-fidelity 3D reconstruction, PointNet++ for robust semantic segmentation, and a customized algorithm for extracting key morphological parameters of tomato seedlings. The proposed framework can be used to reconstruct detailed 3D models at a low computational cost using only ordinary cameras and a limited number of 2D images. We validate the framework on the basis of a tomato seedling dataset and show that our approach outperforms traditional multiview stereo scanners and simple commercial 3D scanners in terms of both detail and efficiency. The accuracy of plant part segmentation reaches 90 %, and the extracted parameters (e.g., leaf area, stem height, branch angle, and internode distance) are highly correlated with the manual measurements (e.g., R 2 = 0.875 for the leaf area). This study provides a low-cost and scalable solution for 3D plant analysis, with direct benefits for automated monitoring of seedling quality in nursery production. Moreover, the proposed framework can be extended to other crops with complex structures, thus supporting wider applications in smart agriculture.
Why it matches plant phenotyping methodsトマト苗の3D再構成、植物部位分割、形態形質抽出を統合した低コスト画像ベース手法を開発・検証しており、植物フェノタイピングが研究の中心である。
abstractwe propose an integrated framework that combines neural radiance fields (NeRFs) for high-fidelity 3D reconstruction, PointNet++ for robust semantic segmentation, and a customized algorithm for extracting key morphological parameters of tomato seedlings.
During the 3D reconstruction of strawberry plants, methods based on 3D Gaussian Splatting (3DGS) face significant challenges due to motion-induced image blur. Such blurring substantially reduces the feature matching accuracy in Structure from Motion (SfM) algorithms and compromises the reliability of camera pose estimation, thereby degrading the quality of subsequent 3DGS reconstruction. This ultimately manifests as geometric distortion and loss of texture details in the reconstructed models. The issue is particularly severe on the surface of strawberry fruits: under blurred image conditions, point cloud registration fails, resulting in the loss of high-frequency details in the high-density achene regions, which blurs seed contours and degrades reconstruction accuracy. To address this technical bottleneck, this study proposes an optimized reconstruction scheme integrated with 3DGS. By incorporating the Canny edge detection algorithm to filter h i gh-quality i n put i m ages, t h e a c curacy of the reconstructed model is significantly improved. The optimized approach achieves remarkable results on the strawberry plant dataset: the average Peak Signal-To-Noise Ratio (PSNR) of the 3DGS model reaches 35.99, representing a 15.2% improvement over the baseline 3DGS. The morphology of high-density achenes on the fruit surface is clearly distinguishable, supporting the accurate monitoring of phenotypic parameters in strawberry plants.
Why it matches plant phenotyping methodsイチゴ植物の3D再構成精度を向上させる画像処理・3DGS手法を開発し、果実表面形態などの表現型パラメータ監視に直接利用するため、方法開発が中心である。
abstractTo address this technical bottleneck, this study proposes an optimized reconstruction scheme integrated with 3DGS.
RiceWheatField / plotMesh / voxelNeRF / 3D Gaussian SplattingLiDAR / point cloudPanicle / ear / spikeLeafRootSeed / grain
Advanced plant phenotyping technologies are vital for trait improvement and accelerating intelligent breeding. Due to the species diversity of plants, existing methods heavily rely on large-scale high-precision manually annotated data. For self-occluded objects at the grain level, unsupervised methods often prove ineffective. This study proposes IPENS, an interactive unsupervised multi-target point cloud extraction method. It utilizes radiance field information to lift 2D masks, segmented by SAM2 (Segment Anything Model 2), into 3D space for target point cloud extraction. A multi-target collaborative optimization strategy addresses the challenge of segmenting multiple targets from a single interaction. On a rice dataset, IPENS achieves a grain-level segmentation mean Intersection over Union (mIoU) of 63.72%. For phenotypic trait estimation, it achieves a grain voxel volume coefficient of determination R 2 = 0.7697 (Root Mean Square Error, RMSE = 0.0025), leaf surface area R 2 = 0.84 (RMSE = 18.93), and leaf length and width prediction accuracies of R 2 = 0.97 and R 2 = 0.87 (RMSE = 1.49 and 0.21). On a wheat dataset, IPENS further improves segmentation performance to a mIoU of 89.68%, with exceptional phenotypic estimation results: panicle voxel volume R 2 = 0.9956 (RMSE = 0.0055), leaf surface area R 2 = 1.00 (RMSE = 0.67), and leaf length and width predictions reaching R 2 = 0.99 and R 2 = 0.92 (RMSE = 0.23 and 0.15). Without requiring annotated data, IPENS rapidly extracts grain-level point clouds for multiple targets within three minutes using single-round image interactions. These features make IPENS a high-quality, non-invasive phenotypic extraction solution for rice and wheat, offering significant potential to enhance intelligent breeding.
Why it matches plant phenotyping methods植物形質抽出のためのNeRF-SAM2融合手法を開発し、作物データセットで分割性能と形質推定精度を検証しているため、方法開発・検証が中心である。
abstractThis study proposes IPENS, an interactive unsupervised multi-target point cloud extraction method.
Reproduction assets foundThe paper's analysis code is publicly available on GitHub. The rice/wheat MMR/MMW phenotype datasets (multi-view images, point clouds, annotations) are only available upon reasonable request, so they are not public.Code · publicCode is available at https://github.com/Vincent-Songwentao/IPENS-Code.git .Open asset ↗https://github.com/Vincent-Songwentao/IPENS-Code.gitlines:472-496Plant phenotyping relevance match · UnverifiedCrossref · OpenAlex · Europe PMC · checked 15 Sept 2026
Plant phenotyping plays a pivotal role in understanding plant traits and their interactions with the environment, making it crucial for advancing precision agriculture and crop improvement. 3D reconstruction technologies have emerged as powerful tools for capturing detailed plant morphology and structure, offering significant potential for accurate and automated phenotyping. This paper provides a comprehensive review of the 3D reconstruction techniques for plant phenotyping, covering classical reconstruction methods, emerging Neural Radiance Fields (NeRF), and the novel 3D Gaussian Splatting (3DGS) approach. Classical methods, which often rely on high-resolution sensors, are widely adopted due to their simplicity and flexibility in representing plant structures. However, they face challenges such as data density, noise, and scalability. NeRF, a recent advancement, enables high-quality, photorealistic 3D reconstructions from sparse viewpoints, but its computational cost and applicability in outdoor environments remain areas of active research. The emerging 3DGS technique introduces a new paradigm in reconstructing plant structures by representing geometry through Gaussian primitives, offering potential benefits in both efficiency and scalability. We review the methodologies, applications, and performance of these approaches in plant phenotyping and discuss their respective strengths, limitations, and future prospects (https://github.com/JiajiaLi04/3D-Reconstruction-Plants). Through this review, we aim to provide insights into how these diverse 3D reconstruction techniques can be effectively leveraged for automated and high-throughput plant phenotyping, contributing to the next generation of agricultural technology.
Why it matches plant phenotyping methods植物フェノタイピング向け3D再構成手法を体系的にレビューし、方法論・応用・性能を扱うことが中心であるため。
abstractThis paper provides a comprehensive review of the 3D reconstruction techniques for plant phenotyping, covering classical reconstruction methods, emerging Neural Radiance Fields (NeRF), and the novel 3D Gaussian Splatting (3DGS) approach.
In-orchard apple size grading remains challenging under occlusions, variable illumination, and irregular fruit morphology. We present a contact–curvature strategy that integrates bending sensors into fin–ray flexible fingers to estimate fruit size during grasping, unifying grasp–and–grade operations. Neural Radiance Fields (NeRF) reconstruction provides offline ground-truth curvature for calibration, yielding strong linear agreement between sensor readings and true curvature (R2=0.9852). Using temporal curvature features across approach, grasp, and steady phases, a gradient-boosting regression model predicts fruit diameter with R2=0.9577 and RMSE = 1.19 mm on the test set. In laboratory conditions, the system achieved an overall grading accuracy of 98.0 % for 200 apples classified into four grades, with a processing capacity of approximately 6 apples·min⁻¹, meeting real-time requirements. In a small-scale orchard pilot study, the system maintainedR2=0.94 andRMSE=1.27 mm, achieving 96 % grading accuracy versus 77 % for a camera-only approach. Compared with vision-only sizing methods, contact-curvature sensing demonstrates inherent robustness to occlusion and illumination while better tolerating morphological irregularities. A methylene–blue protocol confirmed non–destructive operation. Contact–curvature sensing is robust to occlusions/illumination and can, in principle, extend to other near–spherical crops.
Why it matches plant phenotyping methods果実径という植物器官形質を、接触・曲率センサーと回帰モデルで推定する手法を開発し、校正・精度検証・圃場評価まで行っており、表現型取得法が研究の中心である。
abstractWe present a contact–curvature strategy that integrates bending sensors into fin–ray flexible fingers to estimate fruit size during grasping, unifying grasp–and–grade operations.
NeRF / 3D Gaussian SplattingLiDAR / point cloudStem / branch2D/3D reconstructionSegmentationSkeletonization / topologyArchitecture / morphology / geometry
The 3D reconstruction and physical simulation of plants in natural scenes are of significant research and practical value in fields such as agronomy, forestry, ecology, and remote sensing. However, mainstream 3D reconstruction methods generally focus on geometric detail recovery but lack integration with physics-driven approaches, making it challenging to accurately and efficiently simulate the dynamic changes in plant structures. Although the latest physical Gaussian methods can simulate a variety of nonrigid deformations, there is limited consideration of plant-specific structural features, which affects the accuracy of reconstruction and simulation. To address this challenge, an end-to-end woody-plant Gaussian is proposed, which is a framework of high-precision 3D reconstruction and physical simulation for woody plants. This framework begins by fine-tuning a pre-trained plant instance segmentation model tailored for this purpose to reduce environmental noise interference and improve the accuracy of skeleton extraction from point cloud data. It leverages the extracted topology to guide fine-grained hierarchical classification of branches. By segmenting hierarchical radii and cross-sectional proportions, the Gaussian point distribution is constrained, enabling the Gaussian ellipsoids to better align with branch surfaces, thereby enhancing reconstruction details. In the physical simulation stage, the framework incorporates material property variation rules. Using topological guidance, Gaussian ellipsoids are mapped to branch hierarchies, and a cantilever beam physical model predicts Gaussian distributions and covariance matrix parameters. This approach not only improves rendering quality but also enhances the realism of branch-bending simulations. Finally, we evaluate our framework on the photos of real woody plants we took (3D deformable wood plant) and a public dataset (NeRF-synthetic). Compared to existing plant reconstruction methods, woody-plant Gaussian achieves state-of-the-art performance and significantly improves the visual quality of plant physical simulations.
Why it matches plant phenotyping methods木本植物の3D形状・骨格・枝半径を画像から再構成する計算手法を中心に開発しており、植物の構造形質の取得に直接関係する。物理シミュレーションも含む技術評価が行われている。
abstracta framework of high-precision 3D reconstruction and physical simulation for woody plants
Citrus breeding critically relies on precise phenotyping, yet existing RGB/3D phenotyping methods lack standardized workflows balancing affordability, accuracy, and efficiency. An integrated hardware–software pipeline is introduced to address this gap: (1) a custom imaging device for standardized top/side-view capture of the sample fruit, and (2) optimized computer vision algorithms using enhanced 3D Gaussian Splatting (3DGS) to extract 3D structural traits (surface area, volume) and 2D algorithms (e.g., Unet++ and YOLO variants) for 2D trait (width, length, oil cell number, peel/pulp color, peel thickness, and segment number) quantification. Our approach achieved average Pearson correlations above 0.9 between algorithmic and manual measurements across five key traits, confirming measurement precision. The resulting phenotypic profiles revealed biologically significant inter-trait correlations to inform trait-oriented selection decisions and the refinement of breeding strategies. Finally, we operationalized this workflow through user-friendly software, delivering an end-to-end solution that enables high-throughput, low-cost citrus phenotyping with comparatively high accuracy for accelerated breeding applications. Code is available at https://github.com/liangzhao2000/PhenoCitrus . • Low-cost phenotyping device: Dual cameras enable comprehensive fruit imaging. • Precise trait extraction: Multiple deep learning methods extract fruit traits. • Phenotypic profiling: Statistical analysis supports trait-based variety selection.
Why it matches plant phenotyping methods柑橘果実の形態形質を取得・抽出するハードウェア、画像解析、ソフトウェアを開発し、手動測定との相関で検証した中心的なフェノタイピング研究。
abstractAn integrated hardware–software pipeline is introduced to address this gap: (1) a custom imaging device for standardized top/side-view capture of the sample fruit, and (2) optimized computer vision algorithms using enhanced 3D Gaussian Splatting (3DGS) to extract 3D structural traits (surface area, volume) and 2D algorithms (e.g., Unet++ and YOLO variants) for 2D trait (width, length, oil cell number, peel/pulp color, peel thickness, and segment number) quantification.
Rapeseed / canolaNeRF / 3D Gaussian SplattingLiDAR / point cloudFruitCountingSegmentationFruit / seed / panicle traits
Accurate analysis of plant phenotypic traits is crucial for crop breeding and precision agriculture. This study proposes a lightweight semantic segmentation model named KAN-GLNet (Kolmogorov-Arnold Network with Global-Local Feature Modulation), based on an enhanced PointNet++ architecture and integrated with an optimized Density-Based Spatial Clustering of Applications with Noise (DBSCAN) algorithm, to achieve high-precision segmentation and automatic counting of canola siliques. A multi-view point cloud acquisition platform was built, and high-fidelity canola point clouds were reconstructed using Neural Radiance Fields (NeRF) technology. The proposed model includes three key modules: Reverse Bottleneck Kolmogorov-Arnold Network Convolution, a Global-Local Feature Modulation (GLFN) block, and a contrastive learning-based normalization module called ContraNorm. KAN-GLNet contains only 5.72M parameters and achieves 94.50% mIoU, 96.72% mAcc, and 97.77% OAcc in semantic segmentation tasks, outperforming all baseline models. In addition, the DBSCAN workflow was optimized, achieving a counting accuracy of 97.45% in the instance segmentation task. This method achieves an excellent balance between segmentation accuracy and model complexity, providing an efficient solution for high-throughput plant phenotyping. The code and dataset have been made publicly available at: https://anonymous.4open.science/r/KAN-GLNet-6432/.
Why it matches plant phenotyping methodsカノーラ莢のセグメンテーションと自動計数という植物形質抽出手法を、3D点群取得基盤・NeRF再構成・新規モデル・DBSCANワークフローとして開発・評価しており、植物フェノタイピング手法が中心である。
abstractThis study proposes a lightweight semantic segmentation model named KAN-GLNet (Kolmogorov-Arnold Network with Global-Local Feature Modulation), based on an enhanced PointNet++ architecture and integrated with an optimized Density-Based Spatial Clustering of Applications with Noise (DBSCAN) algorithm, to achieve high-precision segmentation and automatic counting of canola siliques.
Reproduction assets foundThe authors explicitly state that their curated code and dataset (canola silique point cloud phenotyping data and KAN-GLNet analysis code) are publicly available at an anonymous.4open.science repository, which is an allowed URL.Code · publicDBSCAN workflow was optimized, achieving a counting accuracy of 97.45% in the instance segmentation task. This method achieves an excellent balance between segmentation accuracy and model complexity, providing an efficient solution for high-throughput plant phenotyping. The code and dataset have been made publicly available at: https://anonymous.4open.science/r/KAN-GLNet-6432/ .
http://dx.doi.org/10.13039/501100001809 National Natural Science Foundation of China
32301762
Liu Jie
This project is supported by National Natural Science Foundation of China, grant number 32301762.
pmc-status-qastatus 0 pmc-status-live yes pmc-status-embargo no pmc-status-released yes pmc-prop-open-access yes pmc-pOpen asset ↗anonymous.4open.science/r/KAN-GLNet-6432lines:1-65Plant phenotyping relevance match · UnverifiedOpenAlex · Crossref · checked 13 Sept 2026
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
Strawberries are among the most economically significant fruits in the United States, generating over $2 billion in annual farm-gate sales and accounting for approximately 13% of the total fruit production value. Plant phenotyping plays a vital role in selecting superior cultivars by characterizing plant traits such as morphology, canopy structure, and growth dynamics. However, traditional plant phenotyping methods are time-consuming, labor-intensive, and often destructive. Recently, neural rendering techniques, notably Neural Radiance Fields (NeRF) and 3D Gaussian Splatting (3DGS), have emerged as powerful frameworks for high-fidelity 3D reconstruction. By capturing a sequence of multi-view images or videos around a target plant, these methods enable non-destructive reconstruction of complex plant architectures. Despite their promise, most current applications of 3DGS in agricultural domains reconstruct the entire scene, including background elements, which introduces noise, increases computational costs, and complicates downstream trait analysis. To address this limitation, we propose a novel object-centric 3D reconstruction framework incorporating a preprocessing pipeline that leverages the Segment Anything Model v2 (SAM-2) and alpha channel background masking to achieve clean strawberry plant reconstructions. This approach produces more accurate geometric representations while substantially reducing computational time. With a background-free reconstruction, our algorithm can automatically estimate important plant traits, such as plant height and canopy width, using DBSCAN clustering and Principal Component Analysis (PCA). Experimental results show that our method outperforms conventional pipelines in both accuracy and efficiency, offering a scalable and non-destructive solution for strawberry plant phenotyping.
Why it matches plant phenotyping methodsイチゴ植物の3D再構成、背景除去、形質推定を一体化した新規フェノタイピング手法の開発・評価が中心であり、植物高と樹冠幅を自動推定して精度・効率を比較している。
abstractwe propose a novel object-centric 3D reconstruction framework incorporating a preprocessing pipeline that leverages the Segment Anything Model v2 (SAM-2) and alpha channel background masking to achieve clean strawberry plant reconstructions.
Plant phenotyping relevance match · UnverifiedOpenAlex · Europe PMC · checked 14 Sept 2026
The 3D heterogeneity in nitrogen content and temperature within the canopy affects canopy photosynthesis. Currently, there are no methods for efficiently assessing the heterogeneous 3D-distribution of leaf nitrogen content and leaf temperature and integrating that information into a 3D model of canopy photosynthesis. We therefore developed a high-throughput pipeline for collecting canopy photosynthesis parameters in maize (Zea mays) by combining several innovations. First, we used readily obtained SPAD502Plus meter readings to infer local leaf nitrogen content. Second, a Bayesian inference method allowed us to parameterize a C4 leaf photosynthesis model. Third, we used neural radiance fields (NeRFs) to recreate 3D plant architecture and SPAD distribution. Finally, we developed an indoor ray tracing and energy balance model to estimate local light distribution and leaf temperature within a canopy. SPAD values showed a distinct 3D pattern, suggesting within-canopy variation in photosynthesis. Bayesian inference efficiently parameterized the C4 leaf photosynthesis model, with estimated parameter values correlating well with SPAD values. In addition, NeRF more accurately reconstructed 3D architecture and estimated 3D SPAD distribution than traditional methods. This resulted in calculated leaf temperatures being similar to measured values. Different model assumptions can cause significant differences in simulated canopy photosynthetic rate. Omitting 3D SPAD heterogeneity alone produced a 1% to 8% difference in simulated canopy photosynthetic rate. Ignoring leaf temperature heterogeneity led to a difference in the calculated canopy photosynthetic rate of only 1% to 3% near the optimal temperature, but of up to 38% at 35 °C. This pipeline can be realized by high-throughput phenotyping platforms, making it suitable for exploring genetic differences and optimizing ideotype design for improved canopy photosynthesis.
Why it matches plant phenotyping methodsトウモロコシ群落の葉窒素・温度・光合成関連形質を3Dで取得・推定する高スループットパイプラインを開発しており、表現型取得手法が研究の中心である。
abstractWe therefore developed a high-throughput pipeline for collecting canopy photosynthesis parameters in maize (Zea mays) by combining several innovations.
NeRF / 3D Gaussian SplattingStereoLeafStem / branchObject detection2D/3D reconstructionSegmentation
Commercial plant phenotyping systems using fixed cameras cannot perceive many plant details due to leaf occlusion. In this paper, we present Botany-Bot, a system for building detailed "annotated digital twins" of living plants using two stereo cameras, a digital turntable inside a lightbox, an industrial robot arm, and 3D segmentated Gaussian Splat models. We also present robot algorithms for manipulating leaves to take high-resolution indexable images of occluded details such as stem buds and the underside/topside of leaves. Results from experiments suggest that Botany-Bot can segment leaves with 90.8% accuracy, detect leaves with 86.2% accuracy, lift/push leaves with 77.9% accuracy, and take detailed overside/underside images with 77.3% accuracy. Code, videos, and datasets are available at https://berkeleyautomation.github.io/Botany-Bot/.
Why it matches plant phenotyping methods植物の遮蔽部位を撮像・セグメンテーションし、3Dデジタルツインとして植物構造を取得するロボット型表現型計測システムの開発が中心である。
abstractwe present Botany-Bot, a system for building detailed "annotated digital twins" of living plants using two stereo cameras, a digital turntable inside a lightbox, an industrial robot arm, and 3D segmentated Gaussian Splat models.
Methods based on upward canopy gap fractions are widely employed to measure in-situ effective LAI (Le) as an alternative to destructive sampling. However, these measurements are limited to point-level and are not practical for scaling up to larger areas. To address the point-to-landscape gap, this study introduces an innovative approach, named NeRF-LAI, for corn and soybean Le estimation that combines gap-fraction theory with the neural radiance field (NeRF) technology, an emerging neural network-based method for implicitly representing 3D scenes using multi-angle 2D images. The trained NeRF-LAI can render downward photorealistic hemispherical depth images from an arbitrary viewpoint in the 3D scene, and then calculate gap fractions to estimate Le. To investigate the intrinsic difference between upward and downward gaps estimations, initial tests on virtual corn fields demonstrated that the downward Le matches well with the upward Le, and the viewpoint height is insensitive to Le estimation for a homogeneous field. Furthermore, we conducted intensive real-world experiments at controlled plots and farmer-managed fields to test the effectiveness and transferability of NeRF-LAI in real-world scenarios, where multi-angle UAV oblique images from different phenological stages were collected for corn and soybeans. Results showed the NeRF-LAI is able to render photorealistic synthetic images with an average peak signal-to-noise ratio (PSNR) of 18.94 for the controlled corn plots and 19.10 for the controlled soybean plots. We further explored three methods to estimate Le from calculated gap fractions: the 57.5° method, the five-ring-based method, and the cell-based method. Among these, the cell-based method achieved the best performance, with the r² ranging from 0.674 to 0.780 and RRMSE ranging from 1.95 % to 5.58 %. The Le estimates are sensitive to viewpoint height in heterogeneous fields due to the difference in the observable foliage volume, but they exhibit less sensitivity to relatively homogeneous fields. Additionally, the cross-site testing for pixel-level LAI mapping showed the NeRF-LAI significantly outperforms the VI-based models, with a small variation of RMSE (0.71 to 0.95 m²/m²) for spatial resolution from 0.5 m to 2.0 m. This study extends the application of gap fraction-based Le estimation from a discrete point scale to a continuous field scale by leveraging implicit 3D neural representations learned by NeRF. The NeRF-LAI method can map Le from raw multi-angle 2D images without prior information, offering a potential alternative to the traditional in-situ plant canopy analyzer with a more flexible and efficient solution.
Why it matches plant phenotyping methodsNeRFとUAV多視点画像を組み合わせ、トウモロコシ・ダイズの葉面積指数を推定・マッピングする手法を開発し、実圃場で性能検証しているため、植物表現型取得が中心である。
abstractthis study introduces an innovative approach, named NeRF-LAI, for corn and soybean Le estimation that combines gap-fraction theory with the neural radiance field (NeRF) technology
Efficient acquisition of 3D plant structures is crucial for investigating growth mechanisms and phenotype analysis. Traditional 3D reconstruction methods exhibited significant limitations when faced with complex background interference, leading to low reconstruction efficiency and compromised result integrity. To address these challenges, a cross-scene 3D plant reconstruction framework P3DFusion was proposed with two key technological modules: (1) GSAM2 multi-view image processing method with Vision Foundation Models, which combines Grounding DINO and Segment Anything Model 2 (SAM2) to achieve high-precision plant segmentation under zero-shot conditions; (2) High-fidelity modeling based on 3D Gaussian splatting (3DGS) to generate high-quality, measurable meshes optimized for plant structural analysis. We evaluated P3DFusion using two datasets: Dataset1 (greenhouse-potted plants) and Dataset2 (open-field sugar beets). The P3DFusion exhibited significant improvements in reconstruction efficiency (SfM-Time reductions of 8.5 %/47.9 %, Total processing time reductions of 60.9 %/65.2 %) and quality metrics (SSIM increases of 12.9 %/19.8 %, PSNR increases of 11.8 %/13 % reaching 24.26 dB/24.75 dB, and LPIPS reductions of 70 %/85 %) for Dataset 1 and Dataset 2, respectively, compared to the original 3DGS. The P3DFusion outperforms InstantNGP (PSNR: 22.3 %/32.9 % increase) and COLMAP (PSNR: 231.4 %/266.1 % increase). Phenotype trait extraction from reconstructed models shows strong consistency with ground truth measurements (R² > 0.93). The proposed method not only provides an effective solution for cross-scene 3D plant reconstruction but also establishes a robust technical foundation for advanced plant phenotype research.
Why it matches plant phenotyping methods植物の3D再構成・セグメンテーションと形質抽出を中核とする手法開発および比較検証であり、植物フェノタイピング手法として明確に該当する。
abstractEfficient acquisition of 3D plant structures is crucial for investigating growth mechanisms and phenotype analysis.
Accurate non-destructive measurement of phenotypic data for leafy vegetable is critical for effective management of breeding, growth monitoring, and yield estimation. However, the scarcity of large-scale stereo matching datasets in agricultural scenarios and the high cost of depth-sensing devices tailored for small targets pose significant challenges to achieving accurate phenotyping of crops in complex growing environments. This study introduces a novel 3D Gaussian Splatting (3DGS)-supervised stereo matching network to enable accurate depth estimation of leafy vegetable. We enhance Gaussian rendering techniques to generate stereo training data from multi-view images. A multi-scale feature extraction network with deformable convolutions is proposed to adaptively capture diverse features, ranging from low-level details (edges and textures) to structural and semantic information. To address the multi-peak characteristics of disparity in leafy vegetable, a top- k algorithm is employed for disparity regression, significantly improving disparity estimation accuracy. To enhance the adaptability of cost aggregation, we construct a combined cost volume through a concatenation and group-wise correlations and leverage feature attention blocks to refine disparity prediction. Additionally, depthwise separable convolutions and optimized methods of the number of iterative blocks for leafy vegetable small targets are introduced to reduce computational complexity and accelerate training and inference. Experimental results demonstrate the effectiveness of our depth estimation method, achieving an Endpoint Error (EPE) of 2.32 pixels overall and 0.91 pixels specifically in leafy vegetable area compared to disparity ground truth. Furthermore, we developed a 3D point cloud phenotyping measurement method using instance segmentation and 3D reconstruction. Notably, the measured height, number of leaves, and surface area achieve R 2 values of 0.9946, 0.9577, and 0.9775 against manual measurements, respectively. These high-precision results underscore the practical applicability of our method and provide a theoretical foundation for deployment in agricultural production systems.
Why it matches plant phenotyping methods葉菜の深度推定と3D点群による形質測定法を開発し、手動測定との比較で高さ・葉数・表面積を検証しており、フェノタイピング手法が研究の中心である。
abstractThis study introduces a novel 3D Gaussian Splatting (3DGS)-supervised stereo matching network to enable accurate depth estimation of leafy vegetable.
Fruit growth has long been described using single- or double-sigmoid curves; however, these temporal models cannot fully capture the spatial heterogeneity that ultimately shapes a fruit. Here, we present a three-dimensional analysis pipeline that non-destructively tracks spatial fruit growth dynamics from field-collected imaginary. Surface landmarks were drawn, and video recordings were taken throughout development for three pome fruits, apple ( Malus × domestica ), Japanese pear ( Pyrus pyrifolia ) and European pear ( Pyrus communis ), and two drupe fruits, peach ( Prunus persica ) and Japanese apricot ( Prunus mume ), to track their motion. Using 3D Gaussian Splatting, we successfully reconstructed 3D models of the fruits, and the landmark displacement could be measured with high accuracy, with R 2 ≥ 0.98 when compared to manual recordings. We found a common spatial growth gradient in the longitudinal growth shared in the pomes and drupes of the Rosaceae; proximal (stem-end) regions exhibited more pronounced growth than the distal (stylar) end. An exception was found in European pear ‘Bartlett,’ which showed relatively vigorous growth in the distal region, explaining its distinct shape with expanded distal end. Transverse expansion varied far less than longitudinal expansion, with a possible association with initial fruit morphology. Inter-fruit growth variability peaked in the fastest-growing regions, particularly in the distal area of the European pear, highlighting the link between growth vigor and phenotypic variance. These results provide foundational insights into the developmental dynamics of both pome and drupe fruits of the Rosaceae family, contributing to the optimization of fruit size, shape, and uniformity.
Why it matches plant phenotyping methods3D Gaussian Splattingを用いて果実の3D再構成と空間的成長形質の非破壊計測パイプラインを開発し、手動記録との精度比較で検証しているため、フェノタイピング手法が中心である。
abstractHere, we present a three-dimensional analysis pipeline that non-destructively tracks spatial fruit growth dynamics
Reproduction assets foundThe paper's data availability statement deposits a subset of the generated 3DGS fruit reconstruction models (the paper's phenotyping outputs) on Figshare with a public DOI; additional data is request-only. No author analysis code is explicitly deposited.Dataset · publicFootnotes
Appendix A
Supplementary data to this article can be found online at https://doi.org/10.1016/j.plaphe.2026.100166 .
Appendix A.
Supplementary data
The following is the Supplementary data to this article:
Multimedia component 1
Multimedia component 1
Data availability
A subset of the generated 3D models is available at https://doi.org/10.6084/m9.figshare.30854579 , where the quality of the 3DGS reconstructions and the marking/measurement procedure can be examined. Additional data may be provided upon reasonable request to the corresponding author.
ReferencesOpen asset ↗figshare · 10.6084/m9.figshare.30854579lines:151-171Plant phenotyping relevance match · UnverifiedOpenAlex · checked 15 Sept 2026
• In this study, we take tomato as an example and propose an improved deep learning combined with clustering algorithm for plant point cloud segmentation. First, a multi-temporal tomato point cloud dataset is constructed by combining multi-view image sequences with neural radiation field (NeRF), and preprocessing and labeling are completed; second, single channel attention (SCA) and global feature aggregation module (GFA) are introduced into the PointNet++ model, respectively, to construct the TomatoSegNet model, which improves the tomato dataset's semantic segmentation performance, while the edge filter was added to the DBSCAN algorithm to improve it and enhance the instance segmentation performance of canopy leaves; finally, a total of six phenotypic parameters were extracted based on the segmented organs. The experimental results show that the TomatoSegNet model has an average precision (mP) of 97.82%, an average recall (mR) of 98.62%, an average F1 score (mF1) of 97.97%, an average intersection and merger ratio (mIoU) of 96.84%, and an overall accuracy (OA) of 94.22% in the tomato dataset, which proves that the use of semantic segmentation algorithms feasibility of stem and leaf segmentation; the improved DBSCAN algorithm achieved an instance segmentation accuracy of 96.03% for leaves, which improved the segmentation accuracy of overlapping leaves; the coefficients of determination between the measured and calculated values of the six phenotypic parameters (plant height, stem thickness, leaf inclination, leaf length, leaf width, and leaf area) were 0.983, 0.903, 0.916, 0.962, 0.951, and 0.978. The method proposed in this study realizes the accurate segmentation and extraction of phenotypic parameters from the 3D point cloud of plants, which provides a valuable reference for automated phenotypic analysis of plants.. In plant phenotyping research, accurate organ segmentation and phenotype extraction is the key to accelerate the process of big data analysis and intelligent breeding.In this study, we take tomato as an example and propose an improved deep learning combined with clustering algorithm for plant point cloud segmentation. First, a multi-temporal tomato point cloud dataset is constructed by combining multi-view image sequences with neural radiation field (NeRF), and preprocessing and labeling are completed; second, single channel attention (SCA) and global feature aggregation module (GFA) are introduced into the PointNet++ model, respectively, to construct the TomatoSegNet model, which improves the tomato dataset's semantic segmentation performance, while the edge filter was added to the DBSCAN algorithm to improve it and enhance the instance segmentation performance of canopy leaves; finally, a total of six phenotypic parameters were extracted based on the segmented organs. The experimental results show that the TomatoSegNet model has an average precision (mP) of 97.82%, an average recall (mR) of 98.62%, an average F1 score (mF1) of 97.97%, an average intersection and merger ratio (mIoU) of 96.84%, and an overall accuracy (OA) of 94.22% in the tomato dataset, which proves that the use of semantic segmentation algorithms feasibility of stem and leaf segmentation; the improved DBSCAN algorithm achieved an instance segmentation accuracy of 96.03% for leaves, which improved the segmentation accuracy of overlapping leaves; the coefficients of determination between the measured and calculated values of the six phenotypic parameters (plant height, stem thickness, leaf inclination, leaf length, leaf width, and leaf area) were 0.983, 0.903, 0.916, 0.962, 0.951, and 0.978. The method proposed in this study realizes the accurate segmentation and extraction of phenotypic parameters from the 3D point cloud of plants, which provides a valuable reference for automated phenotypic analysis of plants.
Why it matches plant phenotyping methods3D点群の生成、深層学習・クラスタリングによる器官分割、6種類の植物表現型抽出を中心に開発・検証した研究であり、植物フェノタイピング手法が明確に中核である。
abstractpropose an improved deep learning combined with clustering algorithm for plant point cloud segmentation
StrawberryTomatoNeRF / 3D Gaussian SplattingFruit2D/3D reconstructionSegmentation
DexFruit is a robotic manipulation framework that enables gentle, autonomous handling of fragile fruit and precise evaluation of damage. Many fruits are fragile and prone to bruising, thus requiring humans to manually harvest them with care. In this work, we demonstrate by using optical tactile sensing, autonomous manipulation of fruit with minimal damage can be achieved. We show that our tactile informed diffusion policies outperform baselines in both reduced bruising and pick-and-place success rate across three fruits: strawberries, tomatoes, and blackberries. In addition, we introduce FruitSplat, a novel technique to represent and quantify visual damage in high-resolution 3D representation via 3D Gaussian Splatting (3DGS). Existing metrics for measuring damage lack quantitative rigor or require expensive equipment. With FruitSplat, we distill a 2D strawberry mask as well as a 2D bruise segmentation mask into the 3DGS representation. Furthermore, this representation is modular and general, compatible with any relevant 2D model. Overall, we demonstrate a 92% grasping policy success rate, up to a 20% reduction in visual bruising, and up to an 31% improvement in grasp success rate on challenging fruit compared to our baselines across our three tested fruits. We rigorously evaluate this result with over 630 trials. Please checkout our website at https://dex-fruit.github.io .
Why it matches plant phenotyping methodsFruitSplatは果実の外観損傷・打撲を3D表現として定量化する画像ベースの植物表現型手法であり、手法開発と大規模な技術評価が中心である。
abstractwe introduce FruitSplat, a novel technique to represent and quantify visual damage in high-resolution 3D representation via 3D Gaussian Splatting (3DGS).
This study presents an end-to-end workflow Segment Any Strawberry Plant (SASP) for estimating strawberry canopy volume from multi-view images. The approach utilizes several recent advances in computer vision and 3D reconstruction. First, a Planar-based Gaussian Splatting Reconstruction (PGSR) method is employed to generate high-fidelity 3D point clouds of strawberry plants, offering improved geometric consistency compared to standard 3D Gaussian Splatting. Next, the Segment Any 3D Gaussians (SAGA) framework is adapted with fully automated prompts derived from YOLO (You Only Look Once) detection and color-based prompt selection which can eliminate the need for manual user input in the segmentation process. The resulting point clouds of plant canopies are calculated via concave hull to estimate their volumes. A reference box of known volume is included in the scene as a calibration object, mapping computed volumes from the virtual 3D space into real-world measurements. Experimental evaluations show that the proposed method achieves high segmentation quality and offers volume estimates across multiple plant shapes. This end-to-end pipeline addresses both the labor-intensive nature of manual canopy measurements and the computational complexity of large-scale 3D reconstructions, offering a potential for high-throughput phenotyping and yield prediction in future strawberry cultivation studies.
Why it matches plant phenotyping methodsイチゴ植物の3D画像から樹冠体積を推定するエンドツーエンド手法を開発・評価しており、植物形態形質の取得が研究の中心である。
abstractThis study presents an end-to-end workflow Segment Any Strawberry Plant (SASP) for estimating strawberry canopy volume from multi-view images.
Plant phenotyping relevance match · UnverifiedCrossref · OpenAlex · Europe PMC · checked 15 Sept 2026
Biomass estimation of oilseed rape is crucial for optimizing crop productivity and breeding strategies. While UAV-based imaging has advanced high-throughput phenotyping, current methods often rely on orthophoto images, which struggle with overlapping leaves and incomplete structural information in complex field environments. This study integrates 3D Gaussian Splatting (3DGS) with the Segment Anything Model (SAM) for precise 3D reconstruction and biomass estimation of oilseed rape. UAV multi-view oblique images from 36 angles were used to perform 3D reconstruction, with the SAM module enhancing point cloud segmentation. The segmented point clouds were then converted into point cloud volumes, which were fitted to ground-measured biomass using linear regression. The results showed that 3DGS (7 k and 30 k iterations) provided high accuracy, with peak signal-to-noise ratios (PSNR) of 27.43 and 29.53 and training times of 7 and 49 min, respectively. This performance exceeded that of structure from motion (SfM) and mipmap Neural Radiance Fields (Mip-NeRF), demonstrating superior efficiency. The SAM module achieved high segmentation accuracy, with a mean intersection over union (mIoU) of 0.961 and an F1-score of 0.980. Additionally, a comparison of biomass extraction models found the point cloud volume model to be the most accurate, with an determination coefficient (R²) of 0.976, root mean square error (RMSE) of 2.92 g/plant, and mean absolute percentage error (MAPE) of 6.81 %, outperforming both the plot crop volume and individual crop volume models. This study highlights the potential of combining 3DGS with multi-view UAV imaging for improved biomass phenotyping.
Why it matches plant phenotyping methodsUAV多視点画像、3D再構成、SAMによる分割、体積からのバイオマス推定を統合・比較検証した、植物表現型取得法が研究の中心である。
abstractThis study integrates 3D Gaussian Splatting (3DGS) with the Segment Anything Model (SAM) for precise 3D reconstruction and biomass estimation of oilseed rape.
Three-dimensional (3D) reconstruction is important for obtaining morphological information and making intelligent management decisions for fruit trees. Thus, a method for the 3D reconstruction and parameters extraction of branches based on Neural Radiation Fields (NeRF) was proposed for walnut (Juglans regia L.) trees. This approach combined Structure from Motion (SfM) with NeRF and used multi-view images to reconstruct branches. First, a dataset of multi-view images of walnut trees was built and camera poses were obtained using SfM. Second, WalnutNeRF was optimized by incorporating hash encoding, piecewise sampler and appearance embedding features to address challenges associated with complex outdoor environments and accurately reconstruct branches. A scale recovery method using calibration objects was employed to extract branch parameters. The effectiveness of WalnutNeRF was evaluated by analyzing rendering performance, reconstruction efficiency, point cloud quality, and the accuracy of extracted branch parameters. WalnutNeRF outperformed existing methods in terms of the quality of rendered images and the accuracy of estimated depth, as determined using PSNR, SSIM, LPIPS, and other metrics. WalnutNeRF resulted in a branch reconstruction accuracy of 90.94 %, with a training time that was 9-time faster than that of SfM-MVS. Compared with SfM-MVS, WalnutNeRF decreased reconstruction errors for the main branches, lateral branches, and watershoots by 72 %, 67 %, and 57 %, respectively, and decreased the errors in length by 7.09 %, 4.33 %, and 65.07 %, respectively. Accordingly, WalnutNeRF decreased the reconstruction time, while increasing accuracy, providing robust support for the development of intelligent management applications (e.g., intelligent pruning) for walnut trees.
Why it matches plant phenotyping methodsNeRFとSfMを用いてクルミ枝の3D形状を再構成し、枝パラメータを抽出・精度評価する手法が研究の中心であるため、植物フェノタイピング手法として含める。
abstracta method for the 3D reconstruction and parameters extraction of branches based on Neural Radiation Fields (NeRF) was proposed for walnut (Juglans regia L.) trees.
Accurately reconstructing tree skeletons from multi-view images is challenging. While most existing works use skeletonization from 3D point clouds, thin branches with low-texture contrast often involve multi-view stereo (MVS) to produce noisy and fragmented point clouds, which break branch connectivity. Leveraging the recent development in accurate mask extraction from images, we introduce a mask-guided graph optimization framework that estimates a 3D skeleton directly from multi-view segmentation masks, bypassing the reliance on point cloud quality. In our method, a skeleton is modeled as a graph whose nodes store positions and radii while its adjacency matrix encodes branch connectivity. We use 3D Gaussian splatting (3DGS) to render silhouettes of the graph and directly optimize the nodes and the adjacency matrix to fit given multi-view silhouettes in a differentiable manner. Furthermore, we use a minimum spanning tree (MST) algorithm during the optimization loop to regularize the graph to a tree structure. Experiments on synthetic and real-world plants show consistent improvements in completeness and structural accuracy over existing point-cloud-based and heuristic baseline methods.
Why it matches plant phenotyping methods植物のマルチビュー画像から樹木の3D骨格・枝構造を推定する計算手法を開発し、実植物で既存法と比較検証しているため、植物形態フェノタイピング手法が中心である。
abstractwe introduce a mask-guided graph optimization framework that estimates a 3D skeleton directly from multi-view segmentation masks
Reproduction assets foundThe paper's authors explicitly state their implementation is publicly available on GitHub, which is the paper-specific computational analysis code for the mask-guided tree skeleton extraction method.Code · publicOur implementation is available in the public GitHub repository ( https://github.com/huntorochi/Masks-to-Skeleton , accessed on 18 May 2025).Open asset ↗huntorochi/Masks-to-Skeletonlines:27-41Plant phenotyping relevance match · UnverifiedOpenAlex · arXiv · checked 8 Sept 2026
NeRF / 3D Gaussian SplattingLiDAR / point cloudRGB / grayscaleWhole plant / canopy / plot / fieldOrgan identification2D/3D reconstructionSegmentation
Organ segmentation of plant point clouds is a prerequisite for the high-resolution and accurate extraction of organ-level phenotypic traits. Although the fast development of deep learning has boosted much research on segmentation of plant point clouds, the existing techniques for organ segmentation still face limitations in resolution, segmentation accuracy, and generalizability across various plant species. In this study, we proposed a novel approach called plant segmentation neural radiance fields (PlantSegNeRF), aiming to directly generate high-precision instance point clouds from multi-view RGB image sequences for a wide range of plant species. PlantSegNeRF performed 2D instance segmentation on the multi-view images to generate instance masks for each organ with a corresponding ID. The multi-view instance IDs corresponding to the same plant organ were then matched and refined using a specially designed instance matching module. The instance NeRF was developed to render an implicit scene, containing color, density, semantic and instance information. The implicit scene was ultimately converted into high-precision plant instance point clouds based on the volume density. The results proved that in semantic segmentation of point clouds, PlantSegNeRF outperformed the commonly used methods, demonstrating an average improvement of 16.1%, 18.3%, 17.8%, and 24.2% in precision, recall, F1-score, and IoU compared to the second-best results on structurally complex species. More importantly, PlantSegNeRF exhibited significant advantages in plant point cloud instance segmentation tasks. Across all plant species, it achieved average improvements of 11.7%, 38.2%, 32.2% and 25.3% in mPrec, mRec, mCov, mWCov, respectively. This study extends the organ-level plant phenotyping and provides a high-throughput way to supply high-quality 3D data for the development of large-scale models in plant science.
Why it matches plant phenotyping methods植物器官のマルチビュー画像から高精度な3Dインスタンスポイントクラウドを再構成・分割する手法を開発し、植物フェノタイピングへの応用と性能比較を行っているため、方法が研究の中心です。
abstractwe proposed a novel approach called plant segmentation neural radiance fields (PlantSegNeRF), aiming to directly generate high-precision instance point clouds from multi-view RGB image sequences for a wide range of plant species.
Cotton is an economically important crop cultivated worldwide for textile production. Breeding programs focus on selecting genotypes with favorable traits for high yields. This study introduced 3D Gaussian Splatting (3DGS) to reconstruct high-fidelity three-dimensional (3D) models and developed a segmentation workflow, Cotton3DGaussians, to analyze cotton bolls and extract architectural traits from single plants. Cotton plants were scanned 360° using a smartphone, and photogrammetry was used to estimate camera parameters and reconstruct a sparse point cloud, which was then optimized into a 3DGS model. In Cotton3DGaussians, 2D masks of bolls segmented from four views were mapped to 3D space, and redundant bolls were removed through cross-view clustering. YOLOv11x and a foundation model, segment anything model (SAM), were compared to obtain 2D masks, with YOLOv11x achieving an F1-score 5.9 % higher than SAM. Phenotypic traits such as boll number, volume, plant height, and canopy size were estimated. The 3DGS model exhibited superior rendering quality, achieving a peak signal-to-noise ratio (PSNR) that was 6.91 higher than NeRF. Cotton3DGaussians effectively segmented 3D bolls from multiple views, with mean absolute percentage errors (MAPE) of 9.23 % for boll number, 3.66 % for canopy size, 2.38 % for plant height, and 8.17 % for boll volume compared to LiDAR ground truth. The regression analysis between convex boll volume and boll weight showed a 19.3 % weight error per plant. This study demonstrates the potential of 3DGS for low-cost, high-fidelity 3D modeling, enabling high-resolution phenotyping and advancing cotton breeding programs. The methodology can also be applied to other crops for improved 3D trait measurement research and enhanced productivity.
Why it matches plant phenotyping methods3Dガウシアンスプラッティングと多視点画像・セグメンテーションを組み合わせ、ワタのボールおよび植物体形質を抽出・検証する手法が研究の中心である。
abstractThis study introduced 3D Gaussian Splatting (3DGS) to reconstruct high-fidelity three-dimensional (3D) models and developed a segmentation workflow, Cotton3DGaussians, to analyze cotton bolls and extract architectural traits from single plants.
Abstract Context Urban green spaces play a vital role in enhancing environmental quality and human well-being. However, traditional assessment methods, such as the green view index, primarily quantify green coverage while neglecting vegetation diversity, color richness, and seasonal dynamics, which are critical for urban livability. Objectives This study develops a multi-temporal and multi-perspective analysis framework for urban green space visualization, introducing the Seasonal Species-Specific Plant View Index (S3PVI) to quantify plant coverage at the species level, capturing seasonal changes and visual diversity. Methods The framework integrates computer vision, deep learning, and 3D reconstruction technologies, including structure from motion and 3D Gaussian splatting. To validate the S3PVI, case studies were conducted in Suita City, Japan, analyzing real-world seasonal vegetation patterns and testing the framework in a virtual park environment to assess its applicability in urban design. Results The S3PVI effectively captured species-specific seasonal patterns, with cherry blossoms peaking at 45.61% visibility in spring and maples at 56.78% in autumn. Comparative analysis revealed distinctive vegetation strategies between streets, with Sanshikisaido showing higher seasonal amplitude but lower consistency than Nakayoshido. Virtual simulations confirmed that multi-species schemes optimally balanced seasonal impact with year-round visual stability. Conclusions The S3PVI framework advances urban vegetation assessment by providing species-specific and seasonally dynamic visual data, supporting evidence-based urban planning for ecological sustainability and livability. Potential applications include brownfield redevelopment, virtual park planning, and urban design simulations.
Why it matches plant phenotyping methods植物の種別・季節別被覆を定量化するS3PVIを開発し、深層学習・コンピュータビジョン・3D再構成で検証しており、植物状態の取得・抽出手法が中心である。
abstractThis study develops a multi-temporal and multi-perspective analysis framework for urban green space visualization, introducing the Seasonal Species-Specific Plant View Index (S3PVI) to quantify plant coverage at the species level, capturing seasonal changes and visual diversity.
Field / plotNeRF / 3D Gaussian SplattingWhole plant / canopy / plot / field2D/3D reconstruction
High-resolution, three-dimensional monitoring is increasingly essential for capturing ecological dynamics, yet conventional approaches such as terrestrial laser scanning (TLS) and photogrammetry remain limited by cost, accessibility, and technical barriers. Here, we introduce and evaluate the application of mobile neural radiance field (NeRF) methods for ecological research. Leveraging consumer-grade smartphones and open-source platforms (e.g. Luma AI), we demonstrate that mobile NeRFs can reconstruct detailed 3D structures of vegetation with accuracy comparable to TLS in open-canopy environments. We assess the strengths and limitations of NeRFs across habitat types, showing that while performance declines under occlusion (e.g. dense canopies), these methods excel at capturing understory complexity, making them particularly valuable for savannas, grasslands, and urban systems. We further explore the potential of radiance fields to integrate hyperspectral and robotic data streams, expanding their utility for dynamic ecosystem monitoring. By reducing hardware requirements and broadening participation, mobile NeRFs offer a promising avenue for democratising ecological data collection and advancing scalable environmental surveillance.
Why it matches plant phenotyping methodsモバイルNeRFによる植生の3D構造取得を中心に開発・評価し、TLSとの精度比較や遮蔽条件での性能評価を行っているため、植生構造の画像ベース表現型計測法として収載対象。
abstractHere, we introduce and evaluate the application of mobile neural radiance field (NeRF) methods for ecological research.
Phenotypic traits and phenotypic extraction at the seedling stage of oilseed rape play a crucial role in assessing oilseed rape growth, breeding new varieties and estimating yield. Manual phenotyping not only consumes a lot of labor and time costs, but even the measurement process can cause structural damage to oilseed rape plants. Existing crop phenotype acquisition methods have limitations in terms of throughput and accuracy, which are difficult to meet the demands of phenotype analysis. We propose an oilseed rape segmentation and phenotyping measurement method based on 3D Gaussian splatting with improved PointNet++. The CKG-PointNet++ network is designed to integrate CGLU and FastKAN convolutional modules in the SA layer, and introduce MogaBlock and a self-attention mechanism in the FP layer to enhance local and global feature extraction. Experiments show that the method achieves a 97.70% overall accuracy (OA) and 96.01% mean intersection over union (mIoU) on the oilseed rape point cloud segmentation task. The extracted phenotypic parameters were highly correlated with manual measurements, with leaf length and width, leaf area and leaf inclination R2 of 0.9843, 0.9632, 0.9806 and 0.8890, and RMSE of 0.1621 cm, 0.1546 cm, 0.6892 cm2 and 2.1144°, respectively. This technique provides a feasible solution for high-throughput and rapid measurement of seedling phenotypes in oilseed rape.
Why it matches plant phenotyping methods3D点群再構成・セグメンテーション・特徴抽出法を開発し、油糧ナタネの葉形態形質を手測定と検証しており、フェノタイピング手法が中心である。
abstractWe propose an oilseed rape segmentation and phenotyping measurement method based on 3D Gaussian splatting with improved PointNet++.
This study explores the development and evaluation of high-throughput 3D plant reconstruction and 2D feature detection and segmentation methods for plant phenotyping. A robotic system was employed to collect datasets of individual cucumber plants, utilizing automated mechanisms for efficient, high-throughput data acquisition. Three types of 3D reconstruction methods called Instant-NGP, Nerfacto, and 3D Gaussian Splatting were adopted and compared in terms of rendering quality and speed.Among them, 3D Gaussian Splatting performed the best, achieving PSNR: 25, SSIM: 0.84, LPIPS: 0.20, and also an impressive rendering speed of 6.39 FPS. Novel viewpoint renderings and depth maps further demonstrated its ability to generate accurate and photo-realistic representations of plants. Additionally, rendered images were utilized for training YOLO models to segment plant features into two classes: leaf and fruit. The YOLOv11s model achieved the highest F1 Score (0.932), balancing speed and accuracy. Ultra-view renderings and segmentation provided valuable insights into plant morphology, including leaf and fruit counts, paving the way for scalable, automated phenotyping applications. This study highlights the potential of integrating 3D Gaussian Splatting with advanced segmentation models for precise and efficient plant phenotyping.
Why it matches plant phenotyping methods植物の高スループット3D再構成、画像セグメンテーション、ロボット計測を開発・比較評価し、葉・果実数や形態の抽出に用いており、植物フェノタイピング手法が中心である。
abstractThis study explores the development and evaluation of high-throughput 3D plant reconstruction and 2D feature detection and segmentation methods for plant phenotyping.
Abstract has not been obtained from indexed metadata or an accessible article page.
Why it matches plant phenotyping methodsトウモロコシとダイズの有効葉面積指数（LAI）をUAV画像から推定する新規手法の開発が題名で明示されており、植物形質の取得・推定が中心である。
titleNeRF-LAI: A hybrid method combining neural radiance field and gap-fraction theory for deriving effective leaf area index of corn and soybean using multi-angle UAV images
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
Field / plotNeRF / 3D Gaussian SplattingFruitCounting2D/3D reconstruction
Accurate 3D fruit counting in orchards is challenging due to heavy occlusion, semantic ambiguity between fruits and surrounding structures, and the high computational cost of volumetric reconstruction. Existing pipelines often rely on multi-view 2D segmentation and dense volumetric sampling, which lead to accumulated fusion errors and slow inference. We introduce FruitLangGS, a language-guided 3D fruit counting framework that reconstructs orchard-scale scenes using an adaptive-density Gaussian Splatting pipeline with radius-aware pruning and tile-based rasterization, enabling scalable 3D representation. During inference, compressed CLIP-aligned semantic vectors embedded in each Gaussian are filtered via a dual-threshold cosine similarity mechanism, retrieving Gaussians relevant to target prompts while suppressing common distractors (e.g., foliage), without requiring retraining or image-space masks. The selected Gaussians are then sampled into dense point clouds and clustered geometrically to estimate fruit instances, remaining robust under severe occlusion and viewpoint variation. Experiments on nine different orchard-scale datasets demonstrate that FruitLangGS consistently outperforms existing pipelines in instance counting recall, avoiding multi-view segmentation fusion errors and achieving up to 99.7% recall on Pfuji-Size_Orch2018 orchard dataset. Ablation studies further confirm that language-conditioned semantic embedding and dual-threshold prompt filtering are essential for suppressing distractors and improving counting accuracy under heavy occlusion. Beyond fruit counting, the same framework enables prompt-driven 3D semantic retrieval without retraining, highlighting the potential of language-guided 3D perception for scalable agricultural scene understanding.
Why it matches plant phenotyping methods果実個数という植物器官形質を推定する3D画像解析手法を開発し、複数データセットで性能検証しているため、植物フェノタイピング手法が研究の中心である。
abstractWe introduce FruitLangGS, a language-guided 3D fruit counting framework
Sorghum canopy architecture in field trials is determined by various phenotypic traits, such plant and panicle count, leaf density and angle and panicle morphology, and canopy height. These traits together affect light capture and biomass production as well as conversion of photosynthates to grain yield. Panicle morphology exhibits considerable variation as influenced by genetics, environmental conditions and management practices. This study presents a framework for the 3D reconstruction of sorghum canopies and phenotyping panicle morphology. First, we developed a scalable, low-altitude Unmanned Aerial Vehicle (UAV)-based protocol that leverages videos for efficient data acquisition, combined with Neural Radiance Fields (NeRF)s to generate high-quality 3D point cloud reconstructions of sorghum canopies. Next, a 3D model was built to simulate 3D sorghum canopies to create annotated datasets for training deep learning-based semantic segmentation and panicle detection algorithms. Finally, we propose SegVoteNet, a novel multi-task deep learning model that integrates VoteNet and PointNet++ within a shared backbone architecture. Designed for semantic segmentation and 3D detection on pure point cloud data, SegVoteNet incorporates a voting and sampling module that leverages segmentation results to optimize object proposal generation. SegVoteNet is robust, achieving 0.986 Mean Average Precision (mAP) @ 0.5 Intersection Over Union (IOU) on synthetic datasets, and 0.850 mAP @ 0.5 IOU on real point cloud datasets for sorghum panicle detection, without fine-tuning. This set of pipelines provides a robust scalable method for phenotyping sorghum panicles in field trials in breeding and commercial applications. Further work is developing a capability to estimate grain number per panicle, which would provide breeders with additional phenotypes to select.
Why it matches plant phenotyping methodsUAV・NeRF・点群再構成と深層学習によるソルガム穂形態の取得・推定パイプラインを開発し、実データで性能評価しており、植物フェノタイピング手法が研究の中心である。
abstractThis study presents a framework for the 3D reconstruction of sorghum canopies and phenotyping panicle morphology.
Plant phenotyping relevance match · UnverifiedCrossref · Europe PMC · OpenAlex · checked 15 Sept 2026
Sorghum canopy architecture in field trials is determined by various phenotypic traits, such plant and panicle count, leaf density and angle and panicle morphology, and canopy height. These traits together affect light capture and biomass production as well as conversion of photosynthates to grain yield. Panicle morphology exhibits considerable variation as influenced by genetics, environmental conditions and management practices. This study presents a framework for the 3D reconstruction of sorghum canopies and phenotyping panicle morphology. First, we developed a scalable, low-altitude Unmanned Aerial Vehicle (UAV)-based protocol that leverages videos for efficient data acquisition, combined with Neural Radiance Fields (NeRF)s to generate high-quality 3D point cloud reconstructions of sorghum canopies. Next, a 3D model was built to simulate 3D sorghum canopies to create annotated datasets for training deep learning-based semantic segmentation and panicle detection algorithms. Finally, we propose SegVoteNet, a novel multi-task deep learning model that integrates VoteNet and PointNet++ within a shared backbone architecture. Designed for semantic segmentation and 3D detection on pure point cloud data, SegVoteNet incorporates a voting and sampling module that leverages segmentation results to optimize object proposal generation. SegVoteNet is robust, achieving 0.986 Mean Average Precision (mAP) @ 0.5 Intersection Over Union (IOU) on synthetic datasets, and 0.850 mAP @ 0.5 IOU on real point cloud datasets for sorghum panicle detection, without fine-tuning. This set of pipelines provides a robust scalable method for phenotyping sorghum panicles in field trials in breeding and commercial applications. Further work is developing a capability to estimate grain number per panicle, which would provide breeders with additional phenotypes to select.
Why it matches plant phenotyping methodsUAV、3D点群、NeRF、深層学習を統合したソルガム穂形態の取得・抽出パイプラインを開発し、実データで性能評価しており、植物表現型計測手法が研究の中心である。
abstractThis study presents a framework for the 3D reconstruction of sorghum canopies and phenotyping panicle morphology.
Plant phenotyping relevance match · UnverifiedOpenAlex · Crossref · Europe PMC · checked 6 Sept 2026
Field / plotNeRF / 3D Gaussian SplattingLiDAR / point cloudStereoWhole plant / canopy / plot / fieldObject detection2D/3D reconstructionStress / disease detectionGrowth / development / phenology
In the context of the booming development of precision agriculture and plant phenotyping, plant 3D reconstruction technology has become a research hotspot, with widespread applications in plant growth monitoring, pest and disease detection, and smart agricultural equipment. Given the complex geometric and textural characteristics of plants, traditional 2D image analysis methods are difficult to meet the modeling requirements, highlighting the growing importance of 3D reconstruction technology. This paper reviews active vision techniques (such as structured light, time-of-flight, and laser scanning methods), passive vision techniques (such as stereo vision and structure from motion), and deep learning-based 3D reconstruction methods (such as NeRF, CNN, and 3DGS). These technologies enhance crop analysis accuracy from multiple perspectives, provide strong support for agricultural production, and significantly promote the development of the field of plant research.
Why it matches plant phenotyping methods植物フェノタイピング向けの3次元再構築技術を体系的にレビューしており、画像ベースの形質取得手法が中心的です。
titleA Review of Optical-Based Three-Dimensional Reconstruction and Multi-Source Fusion for Plant Phenotyping.
AppleMangoPeachPearPlumField / plotNeRF / 3D Gaussian SplattingRGB / grayscaleFruitWhole plant / canopy / plot / field
FruitNeRF++: A Generalized Multi-Fruit Counting Method Utilizing Contrastive Learning and Neural Radiance Fields We introduce FruitNeRF++, a novel fruit-counting approach that combines contrastive learning with neural radiance fields to count fruits from unstructured input photographs of orchards. Our work is based on FruitNeRF [6], which employs a neural semantic field combined with a fruit-specific clusteringapproach. The requirement for adaptation for each fruit type limits the applicability of the method, and makes it difficult to use in practice. To lift this limitation, we design a shape-agnostic multi-fruit counting framework, that complements the RGB and semantic data with instance masks predicted by a vision foundation model. The masks are used to encode the identity of each fruit as instance embeddings into a neural instance field. By volumetrically sampling the neural fields, we extract apoint cloud embedded with the instance features, which can be clustered in a fruit-agnostic manner to obtain the fruit count. We evaluate our approach using a synthetic dataset containing apples, plums, lemons, pears, peaches, and mangoes, as well as a real-world benchmark apple dataset. Our results demonstrate that FruitNeRF++ is easier to control and compares favorably to other state-of-the-art methods.
Why it matches plant phenotyping methods果実を対象とした画像ベースの汎用カウント手法を開発し、合成および実データで評価しているため、植物形質（果実数）の取得・推定が研究の中心です。
abstractWe introduce FruitNeRF++, a novel fruit-counting approach that combines contrastive learning with neural radiance fields to count fruits from unstructured input photographs of orchards.
NeRF / 3D Gaussian SplattingLiDAR / point cloudWhole plant / canopy / plot / fieldMorphology / geometry measurement2D/3D reconstruction
Recent years have seen substantial improvements in the ability to generate synthetic 3D objects using AI. However, generating complex 3D objects, such as plants, remains a considerable challenge. Current generative 3D models struggle with plant generation compared to general objects, limiting their usability in plant analysis tools, which require fine detail and accurate geometry. We introduce PlantDreamer, a novel approach to 3D synthetic plant generation, which can achieve greater levels of realism for complex plant geometry and textures than available text-to-3D models. To achieve this, our new generation pipeline leverages a depth ControlNet, fine-tuned Low-Rank Adaptation and an adaptable Gaussian culling algorithm, which directly improve textural realism and geometric integrity of generated 3D plant models. Additionally, PlantDreamer enables both purely synthetic plant generation, by leveraging L-System-generated meshes, and the enhancement of real-world plant point clouds by converting them into 3D Gaussian Splats. We evaluate our approach by comparing its outputs with state-of-the-art text-to-3D models, demonstrating that PlantDreamer outperforms existing methods in producing high-fidelity synthetic plants. Our results indicate that our approach not only advances synthetic plant generation, but also facilitates the upgrading of legacy point cloud datasets, making it a valuable tool for 3D phenotyping applications.
Why it matches plant phenotyping methods植物の高精細3Dモデル生成・実点群のGaussian Splat変換を開発し、3Dフェノタイピングへの利用を明示的に評価しているため、手法が中心である。
abstractWe introduce PlantDreamer, a novel approach to 3D synthetic plant generation
Advanced plant phenotyping technologies play a crucial role in targeted trait improvement and accelerating intelligent breeding. Due to the species diversity of plants, existing methods heavily rely on large-scale high-precision manually annotated data. For self-occluded objects at the grain level, unsupervised methods often prove ineffective. This study proposes IPENS, an interactive unsupervised multi-target point cloud extraction method. The method utilizes radiance field information to lift 2D masks, which are segmented by SAM2 (Segment Anything Model 2), into 3D space for target point cloud extraction. A multi-target collaborative optimization strategy is designed to effectively resolve the single-interaction multi-target segmentation challenge. Experimental validation demonstrates that IPENS achieves a grain-level segmentation accuracy (mIoU) of 63.72% on a rice dataset, with strong phenotypic estimation capabilities: grain volume prediction yields R2 = 0.7697 (RMSE = 0.0025), leaf surface area R2 = 0.84 (RMSE = 18.93), and leaf length and width predictions achieve R2 = 0.97 and 0.87 (RMSE = 1.49 and 0.21). On a wheat dataset,IPENS further improves segmentation accuracy to 89.68% (mIoU), with equally outstanding phenotypic estimation performance: spike volume prediction achieves R2 = 0.9956 (RMSE = 0.0055), leaf surface area R2 = 1.00 (RMSE = 0.67), and leaf length and width predictions reach R2 = 0.99 and 0.92 (RMSE = 0.23 and 0.15). This method provides a non-invasive, high-quality phenotyping extraction solution for rice and wheat. Without requiring annotated data, it rapidly extracts grain-level point clouds within 3 minutes through simple single-round interactions on images for multiple targets, demonstrating significant potential to accelerate intelligent breeding efficiency.
Why it matches plant phenotyping methods植物の3D画像から穀粒・葉・穂の形態形質を抽出する手法を開発し、複数作物データセットで精度検証しているため、植物フェノタイピング手法が中心である。
abstractThis study proposes IPENS, an interactive unsupervised multi-target point cloud extraction method.
Accurate temporal reconstructions of plant growth are essential for plant phenotyping and breeding, yet remain challenging due to complex geometries, occlusions, and non-rigid deformations of plants. We present a novel framework for building temporal digital twins of plants by combining 3D Gaussian Splatting with a robust sample alignment pipeline. Our method begins by reconstructing Gaussian Splats from multi-view camera data, then leverages a two-stage registration approach: coarse alignment through feature-based matching and Fast Global Registration, followed by fine alignment with Iterative Closest Point. This pipeline yields a consistent 4D model of plant development in discrete time steps. We evaluate the approach on data from the Netherlands Plant Eco-phenotyping Center, demonstrating detailed temporal reconstructions of Sequoia and Quinoa species. Videos and Images can be seen at https://berkeleyautomation.github.io/GrowSplat/
Why it matches plant phenotyping methods植物の多視点画像からGaussian Splattingと位置合わせを用いて、成長の時間的な3D/4Dデジタルツインを構築する手法が研究の中心であり、植物フェノタイピングへの応用も明示されている。
abstractWe present a novel framework for building temporal digital twins of plants by combining 3D Gaussian Splatting with a robust sample alignment pipeline.
Plant phenotyping plays a pivotal role in understanding plant traits and their interactions with the environment, making it crucial for advancing precision agriculture and crop improvement. 3D reconstruction technologies have emerged as powerful tools for capturing detailed plant morphology and structure, offering significant potential for accurate and automated phenotyping. This paper provides a comprehensive review of the 3D reconstruction techniques for plant phenotyping, covering classical reconstruction methods, emerging Neural Radiance Fields (NeRF), and the novel 3D Gaussian Splatting (3DGS) approach. Classical methods, which often rely on high-resolution sensors, are widely adopted due to their simplicity and flexibility in representing plant structures. However, they face challenges such as data density, noise, and scalability. NeRF, a recent advancement, enables high-quality, photorealistic 3D reconstructions from sparse viewpoints, but its computational cost and applicability in outdoor environments remain areas of active research. The emerging 3DGS technique introduces a new paradigm in reconstructing plant structures by representing geometry through Gaussian primitives, offering potential benefits in both efficiency and scalability. We review the methodologies, applications, and performance of these approaches in plant phenotyping and discuss their respective strengths, limitations, and future prospects (https://github.com/JiajiaLi04/3D-Reconstruction-Plants). Through this review, we aim to provide insights into how these diverse 3D reconstruction techniques can be effectively leveraged for automated and high-throughput plant phenotyping, contributing to the next generation of agricultural technology.
Why it matches plant phenotyping methods植物フェノタイピング向け3D再構成手法を体系的にレビューし、各手法の性能・限界・応用を論じることが中心である。
abstractThis paper provides a comprehensive review of the 3D reconstruction techniques for plant phenotyping, covering classical reconstruction methods, emerging Neural Radiance Fields (NeRF), and the novel 3D Gaussian Splatting (3DGS) approach.
The accurate and efficient 3D reconstruction of trees is beneficial for urban forest resource assessment and management. Close-range photogrammetry (CRP) is widely used in the 3D model reconstruction of forest scenes. However, in practical forestry applications, challenges such as low reconstruction efficiency and poor reconstruction quality persist. Recently, novel view synthesis (NVS) technology, such as neural radiance fields (NeRF) and 3D Gaussian splatting (3DGS), has shown great potential in the 3D reconstruction of plants using some limited number of images. However, existing research typically focuses on small plants in orchards or individual trees. It remains uncertain whether this technology can be effectively applied in larger, more complex stands or forest scenes. In this study, we collected sequential images of urban forest plots with varying levels of complexity using imaging devices with different resolutions (cameras on smartphones and UAV). These plots included one with sparse, leafless trees and another with dense foliage and more occlusions. We then performed dense reconstruction of forest stands using NeRF and 3DGS methods. The resulting point cloud models were compared with those obtained through photogrammetric reconstruction and laser scanning methods. The results show that compared to photogrammetric method, NVS methods have a significant advantage in reconstruction efficiency. The photogrammetric method is suitable for relatively simple forest stands, as it is less adaptable to complex ones. This results in tree point cloud models with issues such as excessive canopy noise and wrongfully reconstructed trees with duplicated trunks and canopies. In contrast, NeRF is better adapted to more complex forest stands, yielding tree point clouds of the highest quality that offer more detailed trunk and canopy information. However, it can lead to reconstruction errors in the ground area when the input views are limited. The 3DGS method has a relatively poor capability to generate dense point clouds, resulting in models with low point density, particularly with sparse points in the trunk areas, which affects the accuracy of the diameter at breast height (DBH) estimation. Tree height and crown diameter information can be extracted from the point clouds reconstructed by all three methods, with NeRF achieving the highest accuracy in tree height. However, the accuracy of DBH extracted from photogrammetric point clouds is still higher than that from NeRF point clouds. Meanwhile, compared to ground-level smartphone images, tree parameters extracted from reconstruction results of higher-resolution and varied perspectives of drone images are more accurate. These findings confirm that NVS methods have significant application potential for 3D reconstruction of urban forests.
Why it matches plant phenotyping methodsNVS、写真測量、レーザースキャンを比較して森林の3D再構成と樹高・樹冠径・DBHの抽出精度を評価しており、植物形質の取得手法が研究の中心である。
abstractWe then performed dense reconstruction of forest stands using NeRF and 3DGS methods. The resulting point cloud models were compared with those obtained through photogrammetric reconstruction and laser scanning methods.
Three-dimensional models of trees can help simulate forest resource management, field surveys, and urban landscape design. With the advancement of Computer Vision (CV) and laser remote sensing technology, forestry researchers can use images and point cloud data to perform digital modeling. However, modeling leafless tree models that conform to tree growth rules and have effective branching remains a major challenge. This article proposes a method based on 3D Gaussian Splatting (3D GS) to address this issue. Firstly, we compared the reconstruction of the same tree and confirmed the advantages of the 3D GS method in tree 3D reconstruction. Secondly, seven landscape trees were reconstructed using the 3D GS-based method, to verify the effectiveness of the method. Finally, the 3D reconstructed point cloud was used to generate the QSM and extract tree feature parameters to verify the accuracy of the reconstructed model. Our results indicate that this method can effectively reconstruct the structure of real trees, and especially completely reconstruct 3rd-order branches. Meanwhile, the error of the Diameter at Breast Height (DBH) of the model is below 1.59 cm, with a relative error of 3.8–14.6%. This proves that 3D GS effectively solved the problems of inconsistency between tree models and real growth rules, as well as poor branch structure in tree reconstruction models, providing new insights and research directions for the 3D reconstruction and visualization of landscape trees in the leafless stage.
Why it matches plant phenotyping methods3D Gaussian Splattingによる樹木構造の再構成手法を開発・検証し、再構成モデルから枝構造やDBHなどの植物形質を抽出して精度評価しているため、植物フェノタイピング手法が中心である。
abstractThis article proposes a method based on 3D Gaussian Splatting (3D GS) to address this issue.
Plant phenotyping relevance match · UnverifiedOpenAlex · Europe PMC · Crossref · checked 14 Sept 2026
Accurate and efficient 3D reconstruction of trees is beneficial for urban forest resource assessment and management. Close-Range Photogrammetry (CRP) is widely used in 3D model reconstruction of forest scenes. However, in practical forestry applications, challenges such as low reconstruction efficiency and poor reconstruction quality persist. Recently, Novel View Synthesis (NVS) technology such as Neural Radiance Fields (NeRF) and 3D Gaussian Splatting (3DGS) has shown great potential in the 3D reconstruction of plants using some limited number of images. However, existing research typically focuses on small plants in orchards or individual trees. It remains uncertain whether this technology can be effectively applied in larger, more complex stands or forest scenes. In this study, we collected sequential images of urban forest plots with varying levels of complexity using different imaging devices. We then performed dense reconstruction of forest stand using NeRF and 3DGS methods. The resulting point cloud models were compared with those obtained through photogrammetric reconstruction and laser scanning methods. The results show that compared to photogrammetric method, NVS methods have a significant advantage in reconstruction efficiency. Photogrammetric method is less suited to more complex forest stands, resulting in tree point cloud models with issues such as excessive canopy noise, wrongfully reconstructed trees with duplicated trunks and canopies. In contrast, NeRF is better adapted to more complex forest stands, especially in reconstructing canopy regions. However, it can lead to reconstruction errors in the ground area when the input views are limited. The 3DGS method has a relatively poor capability to generate dense point clouds, resulting in models with low point density, particularly with sparse points in the trunk areas, which affects the accuracy of the diameter at breast height (DBH) estimation. Tree height and crown diameter information can be extracted from the point clouds reconstructed by all three methods, with NeRF achieving the highest accuracy in tree height. However, the accuracy of DBH extracted from photogrammetric point clouds is still higher than that from NeRF point clouds. Meanwhile, compared to ground-level smartphone images, tree parameters extracted from reconstruction results of higher-resolution and varied perspectives of drone images are more accurate. These findings suggest that NVS methods have significant potential for 3D reconstruction of urban forests, providing further technical support for forest resource visualization, inventory and management tasks.
Why it matches plant phenotyping methods森林の3D再構成手法を比較・検証し、樹高、樹冠径、胸高直径という個体レベルの植物形質を点群から抽出して精度評価しているため、方法中心の植物フェノタイピング研究である。
abstractThe resulting point cloud models were compared with those obtained through photogrammetric reconstruction and laser scanning methods.
Automated extraction of plant morphological traits is crucial for supporting crop breeding and agricultural management through high-throughput field phenotyping (HTFP). Solutions based on multi-view RGB images are attractive due to their scalability and affordability, enabling volumetric measurements that 2D approaches cannot directly capture. While advanced methods like Neural Radiance Fields (NeRFs) have shown promise, their application has been limited to counting or extracting traits from only a few plants or organs. Furthermore, accurately measuring complex structures like individual wheat heads-essential for studying crop yields-remains particularly challenging due to occlusions and the dense arrangement of crop canopies in field conditions. The recent development of 3D Gaussian Splatting (3DGS) offers a promising alternative for HTFP due to its high-quality reconstructions and explicit point-based representation. In this paper, we present Wheat3DGS, a novel approach that leverages 3DGS and the Segment Anything Model (SAM) for precise 3D instance segmentation and morphological measurement of hundreds of wheat heads automatically, representing the first application of 3DGS to HTFP. We validate the accuracy of wheat head extraction against high-resolution laser scan data, obtaining per-instance mean absolute percentage errors of 15.1%, 18.3%, and 40.2% for length, width, and volume. We provide additional comparisons to NeRF-based approaches and traditional Muti-View Stereo (MVS), demonstrating superior results. Our approach enables rapid, non-destructive measurements of key yield-related traits at scale, with significant implications for accelerating crop breeding and improving our understanding of wheat development.
Why it matches plant phenotyping methods3D Gaussian SplattingとSAMを用いて小麦穂の3Dセグメンテーションと形態形質抽出手法を開発し、レーザースキャン、NeRF、MVSと比較検証しているため、植物フェノタイピング手法が中心である。
abstractwe present Wheat3DGS, a novel approach that leverages 3DGS and the Segment Anything Model (SAM) for precise 3D instance segmentation and morphological measurement of hundreds of wheat heads automatically
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.
Plant phenotyping relevance match · UnverifiedCrossref · Europe PMC · OpenAlex · checked 6 Sept 2026
High-quality 3D reconstruction and accurate 3D organ segmentation of plants are crucial prerequisites for automatically extracting phenotypic traits. In this study, we first extract a dense point cloud from implicit representations, which derives from reconstructing the maize plants in 3D by using the Nerfacto neural radiance field model. Second, we propose a lightweight point cloud segmentation network (PointSegNet) specifically for stem and leaf segmentation. This network includes a Global-Local Set Abstraction (GLSA) module to integrate local and global features and an Edge-Aware Feature Propagation (EAFP) module to enhance edge-awareness. Experimental results show that our PointSegNet achieves impressive performance compared to five other state-of-the-art deep learning networks, reaching 93.73%, 97.25%, 96.21%, and 96.73% in terms of mean Intersection over Union (mIoU), precision, recall, and F1-score, respectively. Even when dealing with tomato and soybean plants, with complex structures, our PointSegNet also achieves the best metrics. Meanwhile, based on the principal component analysis (PCA), we further optimize the method to obtain the parameters such as leaf length and leaf width by using PCA principal vectors. Finally, the maize stem thickness, stem height, leaf length, and leaf width obtained from our measurements are compared with the manual test results, yielding R 2 values of 0.99, 0.84, 0.94, and 0.87, respectively. These results indicate that our method has high accuracy and reliability for phenotypic parameter extraction. This study throughout the entire process from 3D reconstruction of maize plants to point cloud segmentation and phenotypic parameter extraction, provides a reliable and objective method for acquiring plant phenotypic parameters and will boost plant phenotypic development in smart agriculture.
Why it matches plant phenotyping methods3D再構成、茎葉分割、形質抽出を一体化した植物フェノタイピング手法を開発し、他手法および手測定と比較検証している。
abstractwe propose a lightweight point cloud segmentation network (PointSegNet) specifically for stem and leaf segmentation.
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 · Crossref · checked 15 Sept 2026
Accurate, scalable forestry insights are critical for implementing carbon credit-based reforestation initiatives and data-driven ecosystem management. However, existing forest quantification methods face significant challenges: hand measurement is labor-intensive, time-consuming, and difficult to trust; satellite imagery is not accurate enough; and airborne LiDAR remains prohibitively expensive at scale. In this work, we introduce ForestSplat: an accurate and scalable reforestation monitoring, reporting, and verification (MRV) system built from consumer-grade drone footage and 3D Gaussian Splatting. To evaluate the performance of our approach, we map and reconstruct a 200-acre mangrove restoration project in the Jobos Bay National Estuarine Research Reserve. ForestSplat produces an average mean absolute error (MAE) of 0.17 m and mean error (ME) of 0.007 m compared to canopy height maps derived from airborne LiDAR scans, using 100× cheaper hardware. We hope that our proposed framework can support the advancement of accurate and scalable forestry modeling with consumer-grade drones and computer vision, facilitating a new gold standard for reforestation MRV.
Why it matches plant phenotyping methodsドローン映像と3D Gaussian Splattingにより森林キャノピー高を推定する手法を開発し、航空LiDARと比較検証しているため、植物キャノピー形質の取得方法が中心である。
abstractwe introduce ForestSplat: an accurate and scalable reforestation monitoring, reporting, and verification (MRV) system built from consumer-grade drone footage and 3D Gaussian Splatting.
Plant phenotyping relevance match · UnverifiedOpenAlex · Europe PMC · Crossref · checked 8 Sept 2026
Introduction: Applying 3D reconstruction techniques to individual plants has enhanced high-throughput phenotyping and provided accurate data support for developing "digital twins" in the agricultural domain. High costs, slow processing times, intricate workflows, and limited automation often constrain the application of existing 3D reconstruction platforms. Methods: We develop a 3D reconstruction platform for complex plants to overcome these issues. Initially, a video acquisition system is built based on "camera to plant" mode. Then, we extract the keyframes in the videos. After that, Zhang Zhengyou's calibration method and Structure from Motion(SfM)are utilized to estimate the camera parameters. Next, Camera poses estimated from SfM were automatically calibrated using camera imaging trajectories as prior knowledge. Finally, Object-Based NeRF we proposed is utilized for the fine-scale reconstruction of plants. The OB-NeRF algorithm introduced a new ray sampling strategy that improved the efficiency and quality of target plant reconstruction without segmenting the background of images. Furthermore, the precision of the reconstruction was enhanced by optimizing camera poses. An exposure adjustment phase was integrated to improve the algorithm's robustness in uneven lighting conditions. The training process was significantly accelerated through the use of shallow MLP and multi-resolution hash encoding. Lastly, the camera imaging trajectories contributed to the automatic localization of target plants within the scene, enabling the automated extraction of Mesh. Results and discussion: Our pipeline reconstructed high-quality neural radiance fields of the target plant from captured videos in just 250 seconds, enabling the synthesis of novel viewpoint images and the extraction of Mesh. OB-NeRF surpasses NeRF in PSNR evaluation and reduces the reconstruction time from over 10 hours to just 30 Seconds. Compared to Instant-NGP, NeRFacto, and NeuS, OB-NeRF achieves higher reconstruction quality in a shorter reconstruction time. Moreover, Our reconstructed 3D model demonstrated superior texture and geometric fidelity compared to those generated by COLMAP and Kinect-based reconstruction methods. The $R^2$ was 0.9933,0.9881 and 0.9883 for plant height, leaf length, and leaf width, respectively. The MAE was 2.0947, 0.1898, and 0.1199 cm. The 3D reconstruction platform introduced in this study provides a robust foundation for high-throughput phenotyping and the creation of agricultural "digital twins".
Why it matches plant phenotyping methods植物の3D再構成と、そこからの草丈・葉長・葉幅抽出を中心に開発・比較検証した高スループット表現型解析プラットフォームであり、方法が明確に中心的です。
abstractWe develop a 3D reconstruction platform for complex plants to overcome these issues.
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.
Mesh / voxelNeRF / 3D Gaussian SplattingWhole plant / canopy / plot / fieldMorphology / geometry measurement2D/3D reconstructionTrackingVisualization / data managementArchitecture / morphology / geometryGrowth / development / phenology
Observing plants across time and diverse scenes is critical in uncovering plant growth patterns. Classic methods often struggle to observe or measure plants against complex backgrounds and at different growth stages. This highlights the need for a universal approach capable of providing realistic plant visualizations across time and scene. Here, we introduce PlantGaussian, an approach for generating realistic three-dimensional (3D) visualization for plants across time and scenes. It marks one of the first applications of 3D Gaussian splatting techniques in plant science, achieving high-quality visualization across species and growth stages. By integrating the Segment Anything Model (SAM) and tracking algorithms, PlantGaussian overcomes the limitations of classic Gaussian reconstruction techniques in complex planting environments. A new mesh partitioning technique is employed to convert Gaussian rendering results into measurable plant meshes, offering a methodology for accurate 3D plant morphology phenotyping. To support this approach, PlantGaussian dataset is developed, which includes images of four crop species captured under multiple conditions and growth stages. Using only plant image sequences as input, it computes high-fidelity plant visualization models and 3D meshes for 3D plant morphological phenotyping. Visualization results indicate that most plant models achieve a Peak Signal-to-Noise Ratio (PSNR) exceeding 25, outperforming all models including the original 3D Gaussian Splatting and enhanced NeRF. The mesh results indicate an average relative error of 4% between the calculated values and the true measurements. As a generic 3D digital plant model, PlantGaussian will support expansion of plant phenotype databases, ecological research, and remote expert consultations.
Why it matches plant phenotyping methods3D Gaussian splattingと画像解析を統合し、植物画像から測定可能な3Dメッシュと形態形質を抽出する手法を開発・検証しており、データセットも構築しているため。
abstractA new mesh partitioning technique is employed to convert Gaussian rendering results into measurable plant meshes, offering a methodology for accurate 3D plant morphology phenotyping.
Field / plotNeRF / 3D Gaussian SplattingLiDAR / point cloudStem / branchMorphology / geometry measurement2D/3D reconstructionArchitecture / morphology / geometry
Forest mapping provides critical observational data needed to understand the dynamics of forest environments. Notably, tree diameter at breast height (DBH) is a metric used to estimate forest biomass and carbon dioxide (CO2) sequestration. Manual methods of forest mapping are labor intensive and time consuming, a bottleneck for large-scale mapping efforts. Automated mapping relies on acquiring dense forest reconstructions, typically in the form of point clouds. Terrestrial laser scanning (TLS) and mobile laser scanning (MLS) generate point clouds using expensive LiDAR sensing and have been used successfully to estimate tree diameter. Neural radiance fields (NeRFs) are an emergent technology enabling photorealistic, vision-based reconstruction by training a neural network on a sparse set of input views. In this paper, we present a comparison of MLS and NeRF forest reconstructions for the purpose of trunk diameter estimation in a mixed-evergreen Redwood forest. In addition, we propose an improved DBH-estimation method using convex-hull modeling. Using this approach, we achieved 1.68 cm RMSE (2.81%), which consistently outperformed standard cylinder modeling approaches.
Why it matches plant phenotyping methods森林内の樹木DBHという個体形態形質を、NeRF・MLS再構成と凸包モデルで推定し、手法比較と精度評価を行っているため、植物形質取得法が中心です。
abstractIn this paper, we present a comparison of MLS and NeRF forest reconstructions for the purpose of trunk diameter estimation in a mixed-evergreen Redwood forest.
Reproduction assets foundThe authors explicitly state that their forest datasets (SLAM and NeRF reconstructions, imagery) and TreeTool modeling code contributions are freely available on their public GitHub repository, which is listed in the allowed URLs.Code · publicAR-inertial SLAM with regards to DBH estimation accuracy.
• Improved DBH estimation accuracy via a trunk modeling approach using convex-hull
and density-based filtering methods.
• Open-source modeling code and forest datasets, including SLAM and NeRF recon-
structions of a mixed-evergreen Redwood forest, are freely available at
https://github.com/harelab-ucsc/RedwoodNeRF (accessed on 7 January 2025).
2. Theoretical Background
2.1. The SLAM Approach
The SLAM problem can be broken into two tasks: building a map of the environment
and simultaneously estimating the robot’s trajectory within that map. More specifically,Open asset ↗harelab-ucsc/RedwoodNeRFpdf-raw-page:2 lines:1-50Code / dataset availability confirmedOpenAlex · Europe PMC · checked 15 Sept 2026
BACKGROUND: The reconstruction of 3-dimensional (3D) plant models can offer advantages over traditional 2-dimensional approaches by more accurately capturing the complex structure and characteristics of different crops. Conventional 3D reconstruction techniques often produce sparse or noisy representations of plants using software or are expensive to capture in hardware. Recently, view synthesis models have been developed that can generate detailed 3D scenes, and even 3D models, from only RGB images and camera poses. These models offer unparalleled accuracy but are currently data hungry, requiring large numbers of views with very accurate camera calibration. RESULTS: In this study, we present a view synthesis dataset comprising 20 individual wheat plants captured across 6 different time frames over a 15-week growth period. We develop a camera capture system using 2 robotic arms combined with a turntable, controlled by a re-deployable and flexible image capture framework. We trained each plant instance using two recent view synthesis models: 3D Gaussian splatting (3DGS) and neural radiance fields (NeRF). Our results show that both 3DGS and NeRF produce high-fidelity reconstructed images of a plant subject from views not captured in the initial training sets. We also show that these approaches can be used to generate accurate 3D representations of these plants as point clouds, with 0.74-mm and 1.43-mm average accuracy compared with a handheld scanner for 3DGS and NeRF, respectively. CONCLUSION: We believe that these new methods will be transformative in the field of 3D plant phenotyping, plant reconstruction, and active vision. To further this cause, we release all robot configuration and control software, alongside our extensive multiview dataset. We also release all scripts necessary to train both 3DGS and NeRF, all trained models data, and final 3D point cloud representations. Our dataset can be accessed via https://plantimages.nottingham.ac.uk/ or https://https://doi.org/10.5524/102661. Our software can be accessed via https://github.com/Lewis-Stuart-11/3D-Plant-View-Synthesis.
Why it matches plant phenotyping methods3D植物表現型取得のための撮影システム、再構成手法、データセットを開発し、スキャナとの精度比較で検証しているため、方法が中心的である。
abstractWe develop a camera capture system using 2 robotic arms combined with a turntable, controlled by a re-deployable and flexible image capture framework.
Reproduction assets foundThe paper releases its wheat plant multiview image dataset (via plantimages.nottingham.ac.uk and GigaDB DOI 10.5524/102661), its authors' analysis/capture codebase on GitHub (3D-Plant-View-Synthesis), a Software Heritage archive of that code, and a DOME-ML registry annotation. All are paper-specific, public, and have作者Code · publicruction output across all plants. We hope that our study will provide opportunities for researchers exploring new and improved 3D phenotyping algorithms, 3D reconstruction and view synthesis research, and active vision systems.
Availability of Source Code and Requirements
Project name: 3D Plant View Synthesis:
Project homepage: https://github.com/Lewis-Stuart-11/3D-Plant-View-Synthesis [ 13 ]
Operating system(s): Windows, Ubuntu
Programming language: Python (>=3.8)
License: Apache 2.0
Any restrictions to use by nonacademics: None
Our code has also been archived in Software Heritage [ 66 ]. Functionality, such as Robotic View Capturing, 3DGS to Point Cloud, and our UR5 Configs files, are storOpen asset ↗GitHub · Lewis-Stuart-11/3D-Plant-View-Synthesislines:663-695Plant phenotyping relevance match · UnverifiedOpenAlex · checked 15 Sept 2026
Highlights A method was proposed to simulate and quantify canopy light interception based on structurally accurate 3D branch models of the canopy reconstructed using NeRF technology. Light scattering models, ray tracing algorithms, and the Beer–Lambert law were integrated, with full consideration given to the interactions between light rays and canopy structural units, to realistically simulate light propagation and energy attenuation within the canopy. Introduced two novel metrics, Light Interception Ratio (LIR) and Energy Interception Ratio (EIR), to comprehensively evaluate canopy light interception from quantitative and energetic perspectives. This study provides a tool for simulating and quantifying canopy light interception, offering a valuable reference for canopy pruning. ABSTRACT. The light interception capacity of fruit tree canopies is a critical factor that affects photosynthetic efficiency, yield, and fruit quality. Traditional methods, such as field measurements and radiation transfer models, are widely used but often lack high spatial resolution data and are prone to significant errors in complex conditions. To address this, we proposed a 3D reconstruction-based method to simulate and quantify canopy light interception by modeling spatial light distribution within the canopy. First, NeRF technology was used to reconstruct high-resolution 3D models of 12 plum trees, accurately capturing their complex canopy structures. Second, a canopy light model was developed and combined with ray tracing algorithms to simulate light propagation within the canopy and its interactions with branches and leaves. Finally, two core metrics, the Light Interception Ratio (LIR) and Energy Interception Ratio (EIR), were proposed to quantitatively evaluate canopy light interception capacity from the perspectives of quantity and energy. Using this method, we analyzed the differences in light interception capacities among 12 canopies with varying morphologies and evaluated the impact of different pruning strategies on canopy light interception. The results indicate that canopy density, branch and leaf distribution, and morphological characteristics significantly influence light interception capacity, with open shapes and evenly distributed canopies demonstrating superior performance. For instance, an open-canopy structure with evenly spaced foliage achieved an LIR of 33.48% and an EIR of 7.13%, compared to 32.81% and 6.58% in a denser, more compact canopy. This study offers an intuitive tool to simulate and quantify canopy light interception efficiency, providing scientific insights for pruning and optimal canopy design. Keywords: 3D reconstruction, Canopy light interception, Light distribution simulation, Orchard management optimization, Ray tracing.
Why it matches plant phenotyping methods3D再構成と光線追跡を統合し、果樹キャノピーの光 interception を定量化する手法と指標を開発・適用しており、植物状態の取得・推定が研究の中心である。
abstracttwo core metrics, the Light Interception Ratio (LIR) and Energy Interception Ratio (EIR), were proposed to quantitatively evaluate canopy light interception capacity
Efficient and realistic 3D tree modeling is an important part of low-altitude remote sensing. Trees have high geometric complexity, and the reconstruction effect of traditional modeling methods is not satisfactory. The 3D Gaussian Splatting (3DGS) method is expected to achieve good results in 3D reconstruction of trees at a low cost. A new method based on 3DGS, SAGStree, is proposed for 3D tree componentization model, which mainly realizes the segmentation and reconstruction of tree trunk, branch, and leaves. Specifically, to suppress high-frequency artifacts and enhance the expression of foreground features, a semantic-guided smoothing filter is introduced in the 3DGS training process. Meanwhile, to improve the segmentation accuracy of each tree components, the Multi-scale Fourier Attention Aggregation Network (MFAANet) is constructed, which includes a Multiscale Adaptive Fourier module, a Pyramid Attention Aggregation module, and a Dual-Path Decoder module. In addition, a multi-view mask voting mechanism is designed to achieve accurate reconstruction of complex tree componentization structures through staged recognition and priority label allocation. This method performs well in multiple task computation evaluations, effectively avoids the problems of detail loss and semantic confusion in traditional modeling, and provides technical support for efficient 3D tree modeling and ecological applications.
Why it matches plant phenotyping methods樹木の幹・枝・葉を分割・再構成する3D画像解析手法自体が研究の中心であり、植物の構造形質を抽出する方法開発に該当する。
abstractA new method based on 3DGS, SAGStree, is proposed for 3D tree componentization model, which mainly realizes the segmentation and reconstruction of tree trunk, branch, and leaves.
The Agricultural Digital Twin Network (ADTN) represents one of the key enabling technologies of Agriculture 4.0. However, the unstructured nature of agricultural scenarios poses significant challenges to the geometric fidelity and scalability of ADTNs, as the morphological diversity and entity uniqueness of crops further reduce the rendering accuracy of virtual entities and increase modeling costs. Existing research lacks low-cost solutions for digital twin generation in unstructured agricultural environments, as implicit-based models struggle to capture global invariant features critical to crop structure and texture. To address these challenges, a crop-based ADTN (CBDTN) framework is designed, which incorporates low-cost image acquisition terminals, edge devices, and cloud platforms, thereby optimizing resource utilization through implicit modeling and minimizing costs in resource constrained scenarios. Furthermore, to improve the accuracy of virtual entity generation, a texture prior feature-enabled neural radiance field (TPFNeRF) is proposed, inspired by the human visual system. It emulates the processing mechanism of the visual cortex to extract consistent prior features from multi-view images, thereby providing robust rendering guidance that enhances perceptual fidelity in 3D reconstruction tasks. The experimental results show that the CBDTN with TPFNeRF exhibits superior performance in generating objects with complex textures and structures. In synthetic dataset, it significantly improves reconstruction quality, achieving 6.52 dB PSNR and 4.3% SSIM improvements over the baseline, and delivers overall performance superior to recent methods such as GFB-NeRF and DiSRNeRF. In real-world scenarios, TPFNeRF effectively reconstructs object geometry and color from fixed viewpoints using only 5 MB of model weights, significantly reducing CBDTN deployment costs.
Why it matches plant phenotyping methods作物の多視点画像から形状・構造・色を3D再構成する手法とプラットフォームが研究の中心であり、植物形態の取得に直接関係するため。
abstracta crop-based ADTN (CBDTN) framework is designed, which incorporates low-cost image acquisition terminals, edge devices, and cloud platforms
WatermelonGreenhouseNeRF / 3D Gaussian SplattingRGB / grayscaleLeafWhole plant / canopy / plot / fieldObject detection2D/3D reconstructionSegmentationGrowth / development / phenology
Crop phenotype detection is a precise way to understand and predict the growth of horticultural seedlings in the smart agriculture era to increase the cost-effectiveness and energy efficiency of agricultural production. Crop phenotype detection requires the consideration of plant stature and agricultural devices, like robots and autonomous vehicles, in smart greenhouse ecosystems. However, collecting the imaging dataset is a challenge facing the deep learning detection of plant phenotype given the dynamic changes among leaves and the temporospatial limits of camara sampling. To address this issue, digital cousin is an improvement on digital twins that can be used to create virtual entities of plants through the creation of dynamic 3D structures and plant attributes using RGB image datasets in a simulation environment, using the principles of the variations and interactions of plants in the physical world. Thus, this work presents a two-phase method to obtain the phenotype of horticultural seedling growth. In the first phase, 3D Gaussian splatting is selected to reconstruct and store the 3D model of the plant with 7000 and 30,000 training rounds, enabling the capture of RGB images and the detection of the phenotypes of the seedlings, overcoming temporal and spatial limitations. In the second phase, an improved YOLOv8 model is created to segment and measure the seedlings, and it is modified by adding the LADH, SPPELAN, and Focaler-ECIoU modules. Compared with the original YOLOv8, the precision of our model is 91%, and the loss metric is lower by approximately 0.24. Moreover, a case study of watermelon seedings is examined, and the results of the 3D reconstruction of the seedlings show that our model outperforms classical segmentation algorithms on the main metrics, achieving a 91.0% mAP50 (B) and a 91.3% mAP50 (M).
Why it matches plant phenotyping methods植物の3D再構成、画像取得、セグメンテーション、計測を組み合わせた生育形質抽出手法の開発と評価が中心であり、単なる生物学的実験ではない。
abstractThus, this work presents a two-phase method to obtain the phenotype of horticultural seedling growth.
Pepper / chilliGreenhouseLaboratory / benchtopNeRF / 3D Gaussian SplattingLiDAR / point cloudFruitMorphology / geometry measurement2D/3D reconstructionImage / point-cloud registrationGrowth / development / phenology
Accurate collection of plant phenotyping is critical to optimising sustainable farming practices in precision agriculture. Traditional phenotyping in controlled laboratory environments, while valuable, falls short in understanding plant growth under real-world conditions. Emerging sensor and digital technologies offer a promising approach for direct phenotyping of plants in farm environments. This study investigates a learning-based phenotyping method using neural implicit surfaces reconstruction to achieve accurate in situ phenotyping of pepper plants in greenhouse environments. To quantitatively evaluate the performance of this method, traditional point cloud registration on 3D scanning data is implemented for comparison. Experimental result shows that NIR (neural implicit surfaces reconstruction) achieves competitive accuracy compared to the 3D scanning method. The mean distance error between the scanner-based method and the NeRF (neural radiance fields)-based method is 0.811 mm. This study shows that the learning-based NeRF method has similar accuracy to the 3D scanning-based method but with greater scalability and faster deployment capabilities.
Why it matches plant phenotyping methods植物の3D表現型を取得するNeRFベース手法を開発し、3Dスキャン法と精度比較・検証しており、表現型取得法が研究の中心である。
abstractThis study investigates a learning-based phenotyping method using neural implicit surfaces reconstruction to achieve accurate in situ phenotyping of pepper plants in greenhouse environments.
The rice panicle traits substantially influence grain yield, making them a primary target for rice phenotyping studies. However, most existing techniques are limited to controlled indoor environments and have difficulty in capturing the rice panicle traits under natural growth conditions. Here, we developed PanicleNeRF, a novel method that enables high-precision and low-cost reconstruction of rice panicle three-dimensional (3D) models in the field based on the video acquired by the smartphone. The proposed method combined the large model Segment Anything Model (SAM) and the small model You Only Look Once version 8 (YOLOv8) to achieve high-precision segmentation of rice panicle images. The neural radiance fields (NeRF) technique was then employed for 3D reconstruction using the images with 2D segmentation. Finally, the resulting point clouds are processed to successfully extract panicle traits. The results show that PanicleNeRF effectively addressed the 2D image segmentation task, achieving a mean F1 score of 86.9% and a mean Intersection over Union (IoU) of 79.8%, with nearly double the boundary overlap (BO) performance compared to YOLOv8. As for point cloud quality, PanicleNeRF significantly outperformed traditional SfM-MVS (structure-from-motion and multi-view stereo) methods, such as COLMAP and Metashape. The panicle length was then accurately extracted with the rRMSE of 2.94% for indica and 1.75% for japonica rice. The panicle volume estimated from 3D point clouds strongly correlated with the grain number ( R 2 = 0.85 for indica and 0.82 for japonica ) and grain mass (0.80 for indica and 0.76 for japonica ). This method provides a low-cost solution for high-throughput in-field phenotyping of rice panicles, accelerating the efficiency of rice breeding.
Why it matches plant phenotyping methodsスマートフォン動画、画像セグメンテーション、NeRFによる3D再構成を統合し、イネ穂の形質を抽出する手法を開発・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 developed PanicleNeRF, a novel method that enables high-precision and low-cost reconstruction of rice panicle three-dimensional (3D) models in the field based on the video acquired by the smartphone.
Plant phenotyping relevance match · UnverifiedCrossref · Europe PMC · OpenAlex · checked 6 Sept 2026
Field / plotLaboratory / benchtopNeRF / 3D Gaussian SplattingPhotogrammetry / SfM / MVSLiDAR / point cloudRGB / grayscaleLeafStem / branchWhole plant / canopy / plot / fieldMorphology / geometry measurement
Precise acquisition of potted plant traits has great theoretical significance and practical value for variety selection and guiding scientific cultivation practices. Although phenotypic analysis using two dimensional(2D) digital images is simple and efficient, leaf occlusion reduces the available phenotype information. To address the current challenge of acquiring sufficient non-destructive information from living potted plants, we proposed a three dimensional (3D) phenotyping pipeline that combines neural radiation field reconstruction with path analysis. An indoor collection system was constructed to obtain multi-view image sequences of potted plants. The structure from motion and neural radiance fields (SFM-NeRF) algorithm was then utilized to reconstruct 3D point clouds, which were subsequently denoised and calibrated. Geometric-feature-based path analysis was employed to separate stems from leaves, and density clustering methods were applied to segment the canopy leaves. Phenotypic parameters of potted plant organs were extracted, including height, stem thickness, leaf length, leaf width, and leaf area, and they were manually measured to obtain the true values. The results showed that the coefficient of determination (R2) values, indicating the correlation between the model traits and the true traits, ranged from 0.89 to 0.98, indicating a strong correlation. The reconstruction quality was good. Additionally, 22 potted plants were selected for exploratory experiments. The results indicated that the method was capable of reconstructing plants of various varieties, and the experiments identified key conditions essential for successful reconstruction. In summary, this study developed a low-cost and robust 3D phenotyping pipeline for the phenotype analysis of potted plants. This proposed pipeline not only meets daily production requirements but also advances the field of phenotype calculation for potted plants.
Why it matches plant phenotyping methods植物の3D形状再構成、器官分離・葉分割、形質抽出を一体化したフェノタイピング手法を開発し、手測定との相関で検証しているため、方法が研究の中心である。
abstractwe proposed a three dimensional (3D) phenotyping pipeline that combines neural radiation field reconstruction with path analysis.
Plant phenotyping relevance match · UnverifiedOpenAlex · Europe PMC · checked 15 Sept 2026
WatermelonGreenhouseNeRF / 3D Gaussian SplattingRGB / grayscaleLeafWhole plant / canopy / plot / fieldObject detection2D/3D reconstructionSegmentationGrowth / development / phenology
Crop phenotype detection is a precision way to understand and predict the growth of horticul-tural Seedling in smart agriculture era, to make the agricultural production more costly and en-ergy efficiency. And it bridges the plant statues and the agricultural devices, like robots and au-tonomous vehicles in smart greenhouse ecosystem, to know each other well. However, the im-aging data set collection is a neckless of deep learning of phenotype detection, as the dynamic coverings among leaves and time-spatial limits of camara sampling. To address this issue, digital cousin is boosting digital twins and virtual entities of plants, and considered to create dynamical 3D structures, attributes and RGB image data sets in a simulation environment, with the princi-ples of varies and interactions in physical world. Thus, this work presents a two-phase method to obtain the phenotype of horticultural seedling growth. In the first phase, 3D Gaussian Splatting is selected to reconstruct and store the 3D model of the plant, enabling to capture RGB images and detect the phenotypes of seedlings transcending temporal and spatial limitations. In the second phase, an improved the YOLOv8 model is created to segment and measure the seedlings, and it is modified by adding modules of the LADH, SPPELAN and Focaler-ECIOU to the original YOLOv8 model. Moreover, a case study of watermelon seeding is explored, and the results show that 3D Gaussian Splatting has good performance in 3D reconstruction of seedlings, and the peak sig-nal-to-noise ratio (PSNR) of the trained models is generally above 24. As for semantic segmenta-tion, compared with the original YOLOv8, the computation of our model decreased by 7.50%, the convergence speed increased by 31.35%.
Why it matches plant phenotyping methods幼 horticultural seedlings の3D再構成、画像取得、セグメンテーション、表現型測定を統合した手法開発が中心であり、植物表現型の抽出方法を直接扱っている。
abstractthis work presents a two-phase method to obtain the phenotype of horticultural seedling growth.
The ability to automatically build 3D digital twins of plants from images has countless applications in agriculture, environmental science, robotics, and other fields. However, current 3D reconstruction methods fail to recover complete shapes of plants due to heavy occlusion and complex geometries. In this work, we present a novel method for 3D modeling of agricultural crops based on optimizing a parametric model of plant morphology via inverse procedural modeling. Our method first estimates depth maps by fitting a neural radiance field and then optimizes a specialized loss to estimate morphological parameters that result in consistent depth renderings. The resulting 3D model is complete and biologically plausible. We validate our method on a dataset of real images of agricultural fields, and demonstrate that the reconstructed canopies can be used for a variety of monitoring and simulation applications.
Why it matches plant phenotyping methods画像から作物の完全な3D形状・キャノピー構造を再構成する手法を開発し、実画像データセットで検証しているため、植物表現型取得が中心的です。
abstractwe present a novel method for 3D modeling of agricultural crops based on optimizing a parametric model of plant morphology via inverse procedural modeling.
Biomass estimation of oilseed rape is crucial for optimizing crop productivity and breeding strategies. While UAV-based imaging has advanced high-throughput phenotyping, current methods often rely on orthophoto images, which struggle with overlapping leaves and incomplete structural information in complex field environments. This study integrates 3D Gaussian Splatting (3DGS) with the Segment Anything Model (SAM) for precise 3D reconstruction and biomass estimation of oilseed rape. UAV multi-view oblique images from 36 angles were used to perform 3D reconstruction, with the SAM module enhancing point cloud segmentation. The segmented point clouds were then converted into point cloud volumes, which were fitted to ground-measured biomass using linear regression. The results showed that 3DGS (7k and 30k iterations) provided high accuracy, with peak signal-to-noise ratios (PSNR) of 27.43 and 29.53 and training times of 7 and 49 minutes, respectively. This performance exceeded that of structure from motion (SfM) and mipmap Neural Radiance Fields (Mip-NeRF), demonstrating superior efficiency. The SAM module achieved high segmentation accuracy, with a mean intersection over union (mIoU) of 0.961 and an F1-score of 0.980. Additionally, a comparison of biomass extraction models found the point cloud volume model to be the most accurate, with an determination coefficient (R2) of 0.976, root mean square error (RMSE) of 2.92 g/plant, and mean absolute percentage error (MAPE) of 6.81%, outperforming both the plot crop volume and individual crop volume models. This study highlights the potential of combining 3DGS with multi-view UAV imaging for improved biomass phenotyping.
Why it matches plant phenotyping methodsUAV多視点画像、3D再構成、SAMによる分割を統合し、ナタネのバイオマスを推定・検証する手法が研究の中心である。
abstractThis study integrates 3D Gaussian Splatting (3DGS) with the Segment Anything Model (SAM) for precise 3D reconstruction and biomass estimation of oilseed rape.
Abstract. Rapid urban growth makes green space management crucial to improve citizens’ well-being. Urban olive trees characterize the Italian landscapes and their culture. This study explores different methodologies for urban tree assessment in this context, using an iPhone 14 Pro Max. These included: 1) its integrated Light Detection and Ranging (LiDAR) sensor using the Recon3D app, 2) its camera with Structure from Motion (SfM) techniques, and 3) its camera for generating 3D models using Neural Radiance Fields (NeRF). Additionally, a professional Mobile Laser Scanner (MLS), was used for comparison. Total height (H), canopy base height (CBH) and canopy volume (CV) measurements were extracted using both CloudCompare and allometric formulas. The main aim of this paper is to compare the 3D models of olive trees obtained from low-cost sensors with those generated from the MLS, which is a more accurate device but comes with significantly higher costs. The results, in terms of RMSE (iPhone LiDAR - H: 0.46 m, CBH: 0.12 m, CV: 15.66 m3; iPhone-SfM - H: 0.95 m, CBH: 0.19 m, CV: 25.85 m3; iPhone-NeRF - H: 1.26 m, CBH: 0.31 m, CV: 33.79 m3), bias and volume differences, reveal that the smartphone, in all the methodologies, tends to underestimate measurements as the size of the trees increases. This is due to the higher MLS range of acquisition. Despite these limitations, low-cost solutions like smartphone-based methods can be a viable alternative given their economic accessibility.
Why it matches plant phenotyping methodsスマートフォンLiDAR・SfM・NeRFによるオリーブ樹の3D形状から樹高、樹冠基部高、樹冠体積を抽出し、MLSと精度比較・検証しており、植物形質取得手法が中心である。
abstractThe main aim of this paper is to compare the 3D models of olive trees obtained from low-cost sensors with those generated from the MLS, which is a more accurate device but comes with significantly higher costs.
Plant phenotyping relevance match · UnverifiedCrossref · Europe PMC · OpenAlex · checked 7 Sept 2026
TomatoGreenhouseNeRF / 3D Gaussian SplattingLiDAR / point cloudRGB / grayscaleFruitLeafStem / branchMorphology / geometry measurement2D/3D reconstruction
Recent advancements in digital phenotypic analysis have revolutionized the morphological analysis of crops, offering new insights into genetic trait expressions. This manuscript presents a novel 3D phenotyping pipeline utilizing the cutting-edge Neural Radiance Fields (NeRF) technology, aimed at overcoming the limitations of traditional 2D imaging methods. Our approach incorporates automated RGB image acquisition through unmanned greenhouse robots, coupled with NeRF technology for dense Point Cloud generation. This facilitates non-destructive, accurate measurements of crop parameters such as node length, leaf area, and fruit volume. Our results, derived from applying this methodology to tomato crops in greenhouse conditions, demonstrate a high correlation with traditional human growth surveys. The manuscript highlights the system’s ability to achieve detailed morphological analysis from limited viewpoint of camera, proving its suitability and practicality for greenhouse environments. The results displayed an R-squared value of 0.973 and a Mean Absolute Percentage Error (MAPE) of 0.089 for inter-node length measurements, while segmented leaf point cloud and reconstructed meshes showed an R-squared value of 0.953 and a MAPE of 0.090 for leaf area measurements. Additionally, segmented tomato fruit analysis yielded an R-squared value of 0.96 and a MAPE of 0.135 for fruit volume measurements. These metrics underscore the precision and reliability of our 3D phenotyping pipeline, making it a highly promising tool for modern agriculture.
Why it matches plant phenotyping methodsNeRFとロボットRGB画像を用いてトマトの形態形質を取得・解析する3Dフェノタイピング手法を開発し、従来測定との精度検証も行っており、方法が研究の中心である。
abstractThis manuscript presents a novel 3D phenotyping pipeline utilizing the cutting-edge Neural Radiance Fields (NeRF) technology
Forest mapping provides critical observational data needed to understand the dynamics of forest environments. Notably, tree diameter at breast height (DBH) is a metric used to estimate forest biomass and carbon dioxide sequestration. Manual methods of forest mapping are labor intensive and time consuming, a bottleneck for large-scale mapping efforts. Automated mapping relies on acquiring dense forest reconstructions, typically in the form of point clouds. Terrestrial laser scanning (TLS) and mobile laser scanning (MLS) generate point clouds using expensive LiDAR sensing, and have been used successfully to estimate tree diameter. Neural radiance fields (NeRFs) are an emergent technology enabling photorealistic, vision-based reconstruction by training a neural network on a sparse set of input views. In this paper, we present a comparison of MLS and NeRF forest reconstructions for the purpose of trunk diameter estimation in a mixed-evergreen Redwood forest. In addition, we propose an improved DBH-estimation method using convex-hull modeling. Using this approach, we achieved 1.68 cm RMSE, which consistently outperformed standard cylinder modeling approaches. Our code contributions and forest datasets are freely available at https://github.com/harelab-ucsc/RedwoodNeRF.
Why it matches plant phenotyping methodsNeRFおよびMLSによる森林再構成から樹木DBHを推定し、凸包モデルによる推定法を提案・比較検証しているため、植物形質取得手法が中心です。
abstractIn this paper, we present a comparison of MLS and NeRF forest reconstructions for the purpose of trunk diameter estimation in a mixed-evergreen Redwood forest.
Reproduction assets foundThe authors explicitly state their code contributions and forest datasets (SLAM and NeRF reconstructions used for DBH estimation) are freely available in a public GitHub repository. The other URLs are a cited third-party tool (NeRFCapture) and a background reference (USDA aerial survey), neither of which is a paper-ownDataset · publicOur code contributions and forest datasets are freely available at https://github.com/harelab-ucsc/RedwoodNeRF .Open asset ↗harelab-ucsc/RedwoodNeRFlines:1-52Plant phenotyping relevance match · UnverifiedarXiv · OpenAlex · checked 15 Sept 2026
Field / plotNeRF / 3D Gaussian SplattingPhotogrammetry / SfM / MVSLiDAR / point cloudStem / branchWhole plant / canopy / plot / fieldMorphology / geometry measurement2D/3D reconstructionVisualization / data managementArchitecture / morphology / geometry
Accurate and efficient 3D reconstruction of trees is crucial for forest resource assessments and management. Close-Range Photogrammetry (CRP) is commonly used for reconstructing forest scenes but faces challenges like low efficiency and poor quality. Recently, Novel View Synthesis (NVS) technologies, including Neural Radiance Fields (NeRF) and 3D Gaussian Splatting (3DGS), have shown promise for 3D plant reconstruction with limited images. However, existing research mainly focuses on small plants in orchards or individual trees, leaving uncertainty regarding their application in larger, complex forest stands. In this study, we collected sequential images of forest plots with varying complexity and performed dense reconstruction using NeRF and 3DGS. The resulting point clouds were compared with those from photogrammetry and laser scanning. Results indicate that NVS methods significantly enhance reconstruction efficiency. Photogrammetry struggles with complex stands, leading to point clouds with excessive canopy noise and incorrectly reconstructed trees, such as duplicated trunks. NeRF, while better for canopy regions, may produce errors in ground areas with limited views. The 3DGS method generates sparser point clouds, particularly in trunk areas, affecting diameter at breast height (DBH) accuracy. All three methods can extract tree height information, with NeRF yielding the highest accuracy; however, photogrammetry remains superior for DBH accuracy. These findings suggest that NVS methods have significant potential for 3D reconstruction of forest stands, offering valuable support for complex forest resource inventory and visualization tasks.
Why it matches plant phenotyping methods森林スタンドの3D再構成手法を比較・評価し、樹高やDBHという個体樹木形質の抽出精度を検証しているため、植物フェノタイピング手法が中心です。
abstractperformed dense reconstruction using NeRF and 3DGS. The resulting point clouds were compared with those from photogrammetry and laser scanning.
Field / plotNeRF / 3D Gaussian SplattingLiDAR / point cloudWhole plant / canopy / plot / fieldMorphology / geometry measurementObject detection2D/3D reconstructionArchitecture / morphology / geometry
We evaluate different Neural Radiance Field (NeRF) techniques for the 3D reconstruction of plants in varied environments, from indoor settings to outdoor fields. Traditional methods usually fail to capture the complex geometric details of plants, which is crucial for phenotyping and breeding studies. We evaluate the reconstruction fidelity of NeRFs in 3 scenarios with increasing complexity and compare the results with the point cloud obtained using light detection and ranging as ground truth. In the most realistic field scenario, the NeRF models achieve a 74.6% F1 score after 30 min of training on the graphics processing unit, highlighting the efficacy of NeRFs for 3D reconstruction in challenging environments. Additionally, we propose an early stopping technique for NeRF training that almost halves the training time while achieving only a reduction of 7.4% in the average F1 score. This optimization process substantially enhances the speed and efficiency of 3D reconstruction using NeRFs. Our findings demonstrate the potential of NeRFs in detailed and realistic 3D plant reconstruction and suggest practical approaches for enhancing the speed and efficiency of NeRFs in the 3D reconstruction process.
Why it matches plant phenotyping methods植物の3D形状再構成にNeRFを適用し、LiDARとの比較で再構成精度を評価するとともに、学習高速化手法も提案しており、表現型取得手法が研究の中心である。
abstractWe evaluate different Neural Radiance Field (NeRF) techniques for the 3D reconstruction of plants in varied environments, from indoor settings to outdoor fields.
The rice panicle traits significantly influence grain yield, making them a primary target for rice phenotyping studies. However, most existing techniques are limited to controlled indoor environments and difficult to capture the rice panicle traits under natural growth conditions. Here, we developed PanicleNeRF, a novel method that enables high-precision and low-cost reconstruction of rice panicle three-dimensional (3D) models in the field using smartphone. The proposed method combined the large model Segment Anything Model (SAM) and the small model You Only Look Once version 8 (YOLOv8) to achieve high-precision segmentation of rice panicle images. The NeRF technique was then employed for 3D reconstruction using the images with 2D segmentation. Finally, the resulting point clouds are processed to successfully extract panicle traits. The results show that PanicleNeRF effectively addressed the 2D image segmentation task, achieving a mean F1 Score of 86.9% and a mean Intersection over Union (IoU) of 79.8%, with nearly double the boundary overlap (BO) performance compared to YOLOv8. As for point cloud quality, PanicleNeRF significantly outperformed traditional SfM-MVS (structure-from-motion and multi-view stereo) methods, such as COLMAP and Metashape. The panicle length was then accurately extracted with the rRMSE of 2.94% for indica and 1.75% for japonica rice. The panicle volume estimated from 3D point clouds strongly correlated with the grain number (R2 = 0.85 for indica and 0.82 for japonica) and grain mass (0.80 for indica and 0.76 for japonica). This method provides a low-cost solution for high-throughput in-field phenotyping of rice panicles, accelerating the efficiency of rice breeding.
Why it matches plant phenotyping methodsスマートフォン画像、セグメンテーション、NeRFによる3D再構成を統合し、イネ穂の形質抽出を開発・検証した中心的な手法研究である。
abstractHere, we developed PanicleNeRF, a novel method that enables high-precision and low-cost reconstruction of rice panicle three-dimensional (3D) models in the field using smartphone.
Robotics paired with computer vision are widely used in precision agriculture. Simulations are critical for safety and performance estimation by verifying their routine in a virtual world before real-world testing and deployment. However, many simulators used in agricultural robots lack photorealism in their virtual worlds compared to the real world. We implemented Unreal Engine 5 (UE5) and the Robot Operating System (ROS) to develop a robot simulator tailored to agricultural tasks and synthetic data generation with RGB, segmentation, and depth images. We designed a method for assigning multiple segmentation labels within a single plant mesh. We experimented with a semi-spherical routine for two robot arms to perform 3D point cloud reconstruction across 10 plant assets. We showed our simulator produces much more accurate segmentation images and reconstruction compared to existing UE5 solutions. We extend our results with Neural Radiance Field (NeRF) reconstructions. The packaged simulator, UE5 project, and ROS package with the Python routine can be found at https://github.com/NCSU-BAE-ARLab/AgriRoboSimUE5.
Why it matches plant phenotyping methods植物の3D再構成・自動アノテーションのためのロボットシミュレータと合成画像生成手法を開発し、精度比較も行っており、表現型取得・抽出基盤が研究の中心である。
titlePhotorealistic Arm Robot Simulation for 3D Plant Reconstruction and Automatic Annotation using Unreal Engine 5
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.
Accurate reconstruction of plant phenotypes plays a key role in optimizing sustainable farming practices in the field of Precision Agriculture (PA). Currently, optical sensor-based approaches dominate the field, but the need for high-fidelity 3D reconstruction of crops and plants in unstructured agricultural environments remains challenging. Recently, a promising development has emerged in the form of Neural Radiance Fields (NeRF), a novel method that utilizes neural density fields. This technology has shown impressive performance in various novel vision synthesis tasks, but has remained relatively unexplored in the agricultural context. In our study, we focus on two fundamental tasks within plant phenotyping: (1) the synthesis of 2D novel-view images and (2) the 3D reconstruction of crop and plant models. We explore the world of NeRF, in particular two state-of-the-art (SOTA) methods: Instant-NGP, which excels in generating high-quality images with impressive training and inference speed, and Instant-NSR, which improves the reconstructed geometry by incorporating the Signed Distance Function (SDF) during training. In particular, we present a novel plant phenotype dataset comprising real plant images from production environments. This dataset is a first-of-its-kind initiative aimed at comprehensively exploring the advantages and limitations of NeRF in agricultural contexts. Our experimental results show that NeRF demonstrates commendable performance in the synthesis of novel-view images and is able to achieve reconstruction results that are competitive with Reality Capture, a leading commercial software for 3D Multi-View Stereo (MVS)-based reconstruction. Moreover, our study also highlights certain drawbacks of NeRF, including relatively slow training speeds, performance limitations in cases of insufficient sampling, and challenges in obtaining geometry quality in complex setups. In conclusion, NeRF introduces a new paradigm in plant phenotyping, providing a powerful tool capable of generating multiple representations, such as multi-view images, point cloud and mesh, from a single process.
Why it matches plant phenotyping methodsNeRFを用いた植物表現型の2D画像合成・3D再構成を中心に開発・評価し、データセットも提示しているため、植物フェノタイピング手法として明確に該当する。
abstractIn our study, we focus on two fundamental tasks within plant phenotyping: (1) the synthesis of 2D novel-view images and (2) the 3D reconstruction of crop and plant models.
Accurate collection of plant phenotyping is critical to optimising sustainable farming practices in precision agriculture. Traditional phenotyping in controlled laboratory environments, while valuable, falls short in understanding plant growth under real-world conditions. Emerging sensor and digital technologies offer a promising approach for direct phenotyping of plants in farm environments. This study investigates a learning-based phenotyping method using the Neural Radiance Field to achieve accurate in-situ phenotyping of pepper plants in greenhouse environments. To quantitatively evaluate the performance of this method, traditional point cloud registration on 3D scanning data is implemented for comparison. Experimental result shows that NeRF(Neural Radiance Fields) achieves competitive accuracy compared to the 3D scanning methods. The mean distance error between the scanner-based method and the NeRF-based method is 0.865mm. This study shows that the learning-based NeRF method achieves similar accuracy to 3D scanning-based methods but with improved scalability and robustness.
Why it matches plant phenotyping methodsNeRFを用いた植物の3D表現・形質取得法を開発し、3Dスキャン法との精度比較で検証しており、フェノタイピング手法が研究の中心である。
abstractThis study investigates a learning-based phenotyping method using the Neural Radiance Field to achieve accurate in-situ phenotyping of pepper plants in greenhouse environments.
Three-dimensional (3D) reconstruction of trees has always been a key task in precision forestry management and research. Due to the complex branch morphological structure of trees themselves and the occlusions from tree stems, branches and foliage, it is difficult to recreate a complete three-dimensional tree model from a two-dimensional image by conventional photogrammetric methods. In this study, based on tree images collected by various cameras in different ways, the Neural Radiance Fields (NeRF) method was used for individual tree dense reconstruction and the exported point cloud models are compared with point clouds derived from photogrammetric reconstruction and laser scanning methods. The results show that the NeRF method performs well in individual tree 3D reconstruction, as it has a higher successful reconstruction rate, better reconstruction in the canopy area and requires less images as input. Compared with the photogrammetric dense reconstruction method, NeRF has significant advantages in reconstruction efficiency and is adaptable to complex scenes, but the generated point cloud tend to be noisy and of low resolution. The accuracy of tree structural parameters (tree height and diameter at breast height) extracted from the photogrammetric point cloud is still higher than those derived from the NeRF point cloud. The results of this study illustrate the great potential of the NeRF method for individual tree reconstruction, and it provides new ideas and research directions for 3D reconstruction and visualization of complex forest scenes.
Why it matches plant phenotyping methodsNeRFによる個体樹木の3D再構築を開発・比較評価し、樹高や胸高直径という植物形質の抽出精度も検証しており、フェノタイピング手法が中心である。
abstractthe Neural Radiance Fields (NeRF) method was used for individual tree dense reconstruction and the exported point cloud models are compared with point clouds derived from photogrammetric reconstruction and laser scanning methods
Field / plotNeRF / 3D Gaussian SplattingLiDAR / point cloudWhole plant / canopy / plot / fieldMorphology / geometry measurement2D/3D reconstructionArchitecture / morphology / geometry
We evaluate different Neural Radiance Fields (NeRFs) techniques for the 3D reconstruction of plants in varied environments, from indoor settings to outdoor fields. Traditional methods usually fail to capture the complex geometric details of plants, which is crucial for phenotyping and breeding studies. We evaluate the reconstruction fidelity of NeRFs in three scenarios with increasing complexity and compare the results with the point cloud obtained using LiDAR as ground truth. In the most realistic field scenario, the NeRF models achieve a 74.6% F1 score after 30 minutes of training on the GPU, highlighting the efficacy of NeRFs for 3D reconstruction in challenging environments. Additionally, we propose an early stopping technique for NeRF training that almost halves the training time while achieving only a reduction of 7.4% in the average F1 score. This optimization process significantly enhances the speed and efficiency of 3D reconstruction using NeRFs. Our findings demonstrate the potential of NeRFs in detailed and realistic 3D plant reconstruction and suggest practical approaches for enhancing the speed and efficiency of NeRFs in the 3D reconstruction process.
Why it matches plant phenotyping methods植物の3D形状再構成を対象にNeRF手法を評価し、LiDARを基準とした精度比較と早期停止による高速化を検証しているため、植物表現型取得法が中心である。
abstractWe evaluate different Neural Radiance Fields (NeRFs) techniques for the 3D reconstruction of plants in varied environments, from indoor settings to outdoor fields.
Field / plotNeRF / 3D Gaussian SplattingLiDAR / point cloudWhole plant / canopy / plot / field2D/3D reconstructionArchitecture / morphology / geometry
We evaluate different Neural Radiance Fields (NeRFs) techniques for reconstructing (3D) plants in varied environments, from indoor settings to outdoor fields. Traditional techniques often struggle to capture the complex details of plants, which is crucial for botanical and agricultural understanding. We evaluate three scenarios with increasing complexity and compare the results with the point cloud obtained using LiDAR as ground truth data. In the most realistic field scenario, the NeRF models achieve a 74.65% F1 score with 30 minutes of training on the GPU, highlighting the efficiency and accuracy of NeRFs in challenging environments. These findings not only demonstrate the potential of NeRF in detailed and realistic 3D plant modeling but also suggest practical approaches for enhancing the speed and efficiency of the 3D reconstruction process.
Why it matches plant phenotyping methodsNeRFによる植物の3D形状再構成を複数条件で評価し、LiDAR点群を用いて精度比較しているため、植物形態の取得・再構成手法の技術検証が中心です。
abstractWe evaluate different Neural Radiance Fields (NeRFs) techniques for reconstructing (3D) plants in varied environments, from indoor settings to outdoor fields.
SoybeanNeRF / 3D Gaussian SplattingPhotogrammetry / SfM / MVSRGB / grayscaleRoot2D/3D reconstructionRoot system architecture
Quantifying 3D phenotypic traits for plant shoots and roots is essential to monitor and evaluate plant growth and development. Multi-view stereo (MVS) is a low-cost and widely used photogrammetry method to build 3D point clouds in many agricultural applications. However, it is challenging to adopt MVS directly to obtain complete 3D structures of fine roots for plants such as soybeans. To address this problem, we propose a data processing pipeline incorporating super-resolution (SR) and 3D Gaussian Splatting (GS) to enhance the resolution of 3D root reconstruction, aiming to recover a highly detailed 3D root structure. To this end, first, multi-view images of a soybean root are collected using an RGB camera; second, SR is used to optimize the resolution of the images; third, the processed images are fed to the algorithm structure from motion to obtain a point cloud; and then, 3D GS is applied to enhance the implicit 3D surface reconstruction; finally, perceptual similarity and peak signal-to-noise ratio (PSNR) are used to evaluate the output quality. The method is expected to obtain a high-fidelity 3D reconstruction of plant roots for soybeans and other crops, assisting in the extraction of comprehensive phenotypic traits to accelerate the selection of new varieties for plant breeding.
Why it matches plant phenotyping methods植物根の3D表現型取得を目的に、超解像・SfM・3D Gaussian Splattingを統合した再構成パイプラインを開発しており、フェノタイピング手法が研究の中心である。
abstractwe propose a data processing pipeline incorporating super-resolution (SR) and 3D Gaussian Splatting (GS) to enhance the resolution of 3D root reconstruction
The real-time simulation of large-scale agricultural operations will offer farmers data-driven and physically consistent decision support, facilitated by predictive digital twins. To construct a predictive digital twin, the initial step involves 3D reconstruction of plant geometry. In this paper, a high-resolution, accurate 3D reconstruction of tomato plants, Tomato-NeRF, is proposed, which is specially used for three-dimensional reconstruction of tomato plants. Our approach used a modular design to integrate ideas from their research paper into Tomato-NeRF. By using hash encoding to map coordinates to trainable feature vectors, we balance quality, memory usage, and performance in NeRF training. The proposal sampler targets key regions for rendering, and customized loss functions are designed to optimize specific tasks. The effectiveness of our approach is demonstrated by the ability to generate high-resolution geometric models from phone camera data. Comparative results show that Tomato-NeRF has significant advantages over Instant-NGP and MipNeRF in the tomato plant reconstruction task. The data acquisition method is simpler and more efficient than other reconstruction methods, providing a practical solution for real-time agricultural simulations.
Why it matches plant phenotyping methodsトマト植物の3D形状を画像から再構成するNeRF手法を開発・比較しており、植物形態の取得が研究の中心である。
abstracta high-resolution, accurate 3D reconstruction of tomato plants, Tomato-NeRF, is proposed
Three-dimensional (3D) reconstruction of trees has always been a key task in precision forestry management and research. Due to the complex branch morphological structure of trees themselves and the occlusions from tree stems, branches and foliage, it is difficult to recreate a complete three-dimensional tree model from a two-dimensional image by conventional photogrammetric methods. In this study, based on tree images collected by various cameras in different ways, the Neural Radiance Fields (NeRF) method was used for individual tree reconstruction and the exported point cloud models are compared with point cloud derived from photogrammetric reconstruction and laser scanning methods. The results show that the NeRF method performs well in individual tree 3D reconstruction, as it has higher successful reconstruction rate, better reconstruction in the canopy area, it requires less amount of images as input. Compared with photogrammetric reconstruction method, NeRF has significant advantages in reconstruction efficiency and is adaptable to complex scenes, but the generated point cloud tends to be noisy and low resolution. The accuracy of tree structural parameters (tree height and diameter at breast height) extracted from the photogrammetric point cloud is still higher than those of derived from the NeRF point cloud. The results of this study illustrate the great potential of NeRF method for individual tree reconstruction, and it provides new ideas and research directions for 3D reconstruction and visualization of complex forest scenes.
Why it matches plant phenotyping methodsNeRFによる個体樹木の3D再構成を中心に、写真測量・レーザースキャンと比較検証し、樹高や胸高直径という植物形質を抽出しているため。
abstractthe Neural Radiance Fields (NeRF) method was used for individual tree reconstruction and the exported point cloud models are compared with point cloud derived from photogrammetric reconstruction and laser scanning methods
Accurate reconstruction of plant phenotypes plays a key role in optimising sustainable farming practices in the field of Precision Agriculture (PA). Currently, optical sensor-based approaches dominate the field, but the need for high-fidelity 3D reconstruction of crops and plants in unstructured agricultural environments remains challenging. Recently, a promising development has emerged in the form of Neural Radiance Field (NeRF), a novel method that utilises neural density fields. This technique has shown impressive performance in various novel vision synthesis tasks, but has remained relatively unexplored in the agricultural context. In our study, we focus on two fundamental tasks within plant phenotyping: (1) the synthesis of 2D novel-view images and (2) the 3D reconstruction of crop and plant models. We explore the world of neural radiance fields, in particular two SOTA methods: Instant-NGP, which excels in generating high-quality images with impressive training and inference speed, and Instant-NSR, which improves the reconstructed geometry by incorporating the Signed Distance Function (SDF) during training. In particular, we present a novel plant phenotype dataset comprising real plant images from production environments. This dataset is a first-of-its-kind initiative aimed at comprehensively exploring the advantages and limitations of NeRF in agricultural contexts. Our experimental results show that NeRF demonstrates commendable performance in the synthesis of novel-view images and is able to achieve reconstruction results that are competitive with Reality Capture, a leading commercial software for 3D Multi-View Stereo (MVS)-based reconstruction. However, our study also highlights certain drawbacks of NeRF, including relatively slow training speeds, performance limitations in cases of insufficient sampling, and challenges in obtaining geometry quality in complex setups.
Why it matches plant phenotyping methods植物表現型の2D画像合成・3D再構成を対象にNeRF手法を検討し、実植物データセットを提示して既存手法と性能比較しているため、表現型取得・抽出法が中心である。
abstractIn our study, we focus on two fundamental tasks within plant phenotyping: (1) the synthesis of 2D novel-view images and (2) the 3D reconstruction of crop and plant models.
In precision agriculture and plant biology, monitoring nutrient stress in crops is of paramount importance for ensuring optimal yield and resource utilization. However, nutrient stress phenotypes can be nuanced, subtle, and display in a variety of ways. We propose using Neural Radiance Fields (NeRFs) for the organized reconstruction of plant structures to observe the changes of plant structure and color under nutrient stress. Neural Radiance Fields, a cutting-edge technique in computer vision, leverage neural networks to model complex high-frequency geometry directly from 2D images, offering high-fidelity reconstructions. This methodology holds immense potential for plant imaging, as it allows for the creation of detailed and organized 3D models that can capture subtle alterations in plant morphology associated with nutrient stress responses. The proposed methodology involves the acquisition of high-resolution images of plants under different nutrient conditions. These images are inputted to the NeRFStudio Nerfacto implementation, a NeRF model that is a aggregation of many different existing models. A 3D reconstruction of the scene is outputted from the model and can be further reduced to a point cloud containing point locations, colors, and normals. Phenotypic traits are then calculated from the point clouds. The reconstructed plant models enable the quantitative analysis of morphological changes associated with nutrient stress. This includes alterations in leaf size, branching patterns, and overall plant geometry. The utilization of NeRFs allows for non-destructive monitoring, offering a significant advantage over traditional methods that may be labor-intensive or invasive. This research not only contributes to the field of precision agriculture but also presents a powerful tool for plant biologists to deepen their understanding of how nutrient stress impacts plant architecture. The insights gained from this approach have the potential to inform precision nutrient management strategies, leading to more sustainable and efficient agricultural practices.
Why it matches plant phenotyping methodsNeRFによる植物の3D再構成と点群からの形態形質抽出が研究の中心であり、栄養ストレスに伴う葉サイズ・分枝・植物形状を定量化する手法を提案している。
abstractWe propose using Neural Radiance Fields (NeRFs) for the organized reconstruction of plant structures to observe the changes of plant structure and color under nutrient stress.
In precision agriculture and plant biology, monitoring nutrient stress in crops is of paramount importance for ensuring optimal yield and resource utilization. However, nutrient stress phenotypes can be nuanced, subtle, and display in a variety of ways. We propose using Neural Radiance Fields (NeRFs) for the organized reconstruction of plant structures to observe the changes of plant structure and color under nutrient stress.Neural Radiance Fields, a cutting-edge technique in computer vision, leverage neural networks to model complex high-frequency geometry directly from 2D images, offering high-fidelity reconstructions. This methodology holds immense potential for plant imaging, as it allows for the creation of detailed and organized 3D models that can capture subtle alterations in plant morphology associated with nutrient stress responses.The proposed methodology involves the acquisition of high-resolution images of plants under different nutrient conditions. These images are inputted to the NeRFStudio Nerfacto implementation, a NeRF model that is a aggregation of many different existing models. A 3D reconstruction of the scene is outputted from the model and can be further reduced to a point cloud containing point locations, colors, and normals. Phenotypic traits are then calculated from the point clouds. The reconstructed plant models enable the quantitative analysis of morphological changes associated with nutrient stress. This includes alterations in leaf size, branching patterns, and overall plant geometry. The utilization of NeRFs allows for non-destructive monitoring, offering a significant advantage over traditional methods that may be labor-intensive or invasive.This research not only contributes to the field of precision agriculture but also presents a powerful tool for plant biologists to deepen their understanding of how nutrient stress impacts plant architecture. The insights gained from this approach have the potential to inform precision nutrient management strategies, leading to more sustainable and efficient agricultural practices.
Why it matches plant phenotyping methodsNeRFによる植物の3D再構成から形態形質を抽出する手法が研究の中心であり、栄養ストレス表現型の定量化を目的としている。
abstractWe propose using Neural Radiance Fields (NeRFs) for the organized reconstruction of plant structures to observe the changes of plant structure and color under nutrient stress.
In precision agriculture and plant biology, monitoring nutrient stress in crops is of paramount importance for ensuring optimal yield and resource utilization.However, nutrient stress phenotypes can be nuanced, subtle, and display in a variety of ways.We propose using Neural Radiance Fields (NeRFs) for the organized reconstruction of plant structures to observe the changes of plant structure and color under nutrient stress.Neural Radiance Fields, a cutting-edge technique in computer vision, leverage neural networks to model complex high-frequency geometry directly from 2D images, offering high-fidelity reconstructions.This methodology holds immense potential for plant imaging, as it allows for the creation of detailed and organized 3D models that can capture subtle alterations in plant morphology associated with nutrient stress responses.The proposed methodology involves the acquisition of high-resolution images of plants under different nutrient conditions.These images are inputted to the NeRFStudio Nerfacto implementation, a NeRF model that is a aggregation of many different existing models.A 3D reconstruction of the scene is outputted from the model and can be further reduced to a point cloud containing point locations, colors, and normals.Phenotypic traits are then calculated from the point clouds.The reconstructed plant models enable the quantitative analysis of morphological changes associated with nutrient stress.This includes alterations in leaf size, branching patterns, and overall plant geometry.The utilization of NeRFs allows for non-destructive monitoring, offering a significant advantage over traditional methods that may be labor-intensive or invasive.This research not only contributes to the field of precision agriculture but also presents a powerful tool for plant biologists to deepen their understanding of how nutrient stress impacts plant architecture.The insights gained from this approach have the potential to inform precision nutrient management strategies, leading to more sustainable and efficient agricultural practices.
Why it matches plant phenotyping methodsNeRFによる3D植物再構成から形態形質を抽出する手法が研究の中心であり、栄養ストレス下の葉サイズ・分枝・植物形状を定量化している。
abstractWe propose using Neural Radiance Fields (NeRFs) for the organized reconstruction of plant structures to observe the changes of plant structure and color under nutrient stress.