A High-Precision Three-Dimensional Branch Reconstruction Method for Tree Morphological Phenotype
The three-dimensional architecture of the trunk and branch system of individual trees constitutes essential information for tree breeding trait evaluation, the construction of functional-structural plant models (FSPM), and precision forest management; its morphological traits are typically represented by diameter at breast height (DBH), tree height, branch order, and the 3D branching topology. In recent years, point-cloud-based tree structure parsing has advanced rapidly. However, constrained by field acquisition conditions and sensor principles, tree point clouds commonly exhibit missing data and noise, causing locally non-uniform density and introducing noise interference. Under these conditions, traditional QSM (Quantitative Structure Models) tend to produce cross-section distortion and inflated volume, degrading both modeling accuracy and the reliability of parameter extraction; moreover, their radius estimation usually relies on species-specific calibration, limiting cross-species application. To address these limitations, this study proposes a high-precision three-dimensional reconstruction method for the trunk and branch system to extract tree morphological phenotypes, which achieves robust reconstruction under noise and missing data through density-adaptive clustering and biologically mechanistic self-adaptive WBE radius simulation. Experimental results show that, compared with existing mainstream methods, the proposed method yields better modeling performance and stronger robustness on both artificial and real branch point clouds, reaching centimeter-level accuracy.
Authors
- Tingdong Yang
- Xueyan Zhu (ORCID: https://orcid.org/0000-0001-8532-3658)
- Huaiqing Zhang (ORCID: https://orcid.org/0000-0003-3874-5326)
- Yili Zheng (ORCID: https://orcid.org/0000-0002-2240-1396)
- Jie Yang
Institutions
- Beijing Forestry University (CN)
- Institute of Forest Resource Information Techniques (CN)
- Chinese Academy of Forestry (CN)
- State Forestry and Grassland Administration (CN)
Publication Details
- Journal
- Remote Sensing
- Published
- 2026-11-20
- DOI
- https://doi.org/10.3390/rs18223994
- Primary Topic
- Remote Sensing and LiDAR Applications
- Type
- article
- Field-Weighted Citation Impact
- 0.00