Accurate LAI estimation of soybean plants in the field using deep learning and clustering algorithms

Abstract

The leaf area index (LAI) is a critical parameter for characterizing plant foliage abundance, canopy structure changes, and vegetation productivity in ecosystems. Traditional phenological measurements are often destructive, time-consuming, and labor-intensive. This paper proposes a high-throughput 3D point cloud data processing pipeline to segment field soybean plants and estimate their LAI. The 3D point cloud data is obtained from a UAV equipped with a LiDAR camera. First, The PointNet++ model was applied to simplify the segmentation process by isolating field soybean plants from their surroundings and eliminating environmental complexities. Subsequently, individual segmentation was achieved using the Watershed approach and k-means clustering algorithms, segmenting the field soybeans into individual plants. Finally, the LAI of soybean plant was estimated using a machine learning method and validated against measured values. The PointNet++ model improved segmentation accuracy by 6.73%, and the watershed algorithm achieved F1 scores of 0.89-0.90, outperforming k-means in complex adhesion cases. For LAI estimation, the SVM model showed the highest accuracy (R² = 0.79, RMSE = 0.47), with RF and XGBoost also performing well (R² > 0.69, RMSE< 0.65). This indicates that the individual segmentation algorithm, Watershed-based approach combined with PointNet++, can serve as a crucial foundation for extracting high-throughput plant phenotypic data. The experimental results confirm that the proposed method can rapidly calculate the morphological parameters of each soybean plant, making it suitable for high-throughput soybean phenotyping.

Plant phenotyping relevance

UAV-LiDAR点群、深層学習・クラスタリングを用いて個体分割とLAI推定を開発・検証する、植物表現型取得手法が中心の研究。

Code and data availability

The article describes UAV-LiDAR soybean point cloud data, a 126-file annotated dataset, PointNet++/watershed/k-means pipelines, and SVM/RF/XGBoost LAI models, but contains no data or code availability statement, no repository deposit, and no public URL for any paper-specific asset. Only generic tools (CloudCompare, PCL

No evidence-backed public reproduction asset is currently recorded.