Graph-based methods for analyzing orchard tree structure using noisy point cloud data
This work addresses the need for rapid, low-quality data processing in orchard digitization for improving agricultural yield, representing an incremental advance over existing methods.
The paper tackled the problem of analyzing orchard tree structure from noisy LiDAR point clouds, achieving improved tree location and segmentation with F1 scores of 0.774 and 0.915, respectively, while trunk classification had a lower F1 of 0.490 but faster runtime.
Digitisation of fruit trees using LiDAR enables analysis which can be used to better growing practices to improve yield. Sophisticated analysis requires geometric and semantic understanding of the data, including the ability to discern individual trees as well as identifying leafy and structural matter. Extraction of this information should be rapid, as should data capture, so that entire orchards can be processed, but existing methods for classification and segmentation rely on high-quality data or additional data sources like cameras. We present a method for analysis of LiDAR data specifically for individual tree location, segmentation and matter classification, which can operate on low-quality data captured by handheld or mobile LiDAR. Our methods for tree location and segmentation improved on existing methods with an F1 score of 0.774 and a v-measure of 0.915 respectively, while trunk matter classification performed poorly in absolute terms with an average F1 score of 0.490 on real data, though consistently outperformed existing methods and displayed a significantly shorter runtime.