.png){kind=icon}
.png){kind=icon}
Return
SATree: Structure-aware tree instance segmentation from 3D LiDAR point clouds
.png){kind=icon}
PRE
AI
DOI：10.1016/j.ufug.2026.129414.png)
.png){kind=icon}
Original request for help.png){kind=icon}
Share.png){kind=icon}
Save
Abstract
En 中文
• Structure-aware approach for 3D tree instance segmentation in large-scale forestry areas. • Fusion of crucial tree parts, including crowns and stems, to explicitly enhance segmentation robustness against crown overlaps and varying tree shapes. • Integration of heatmap prediction to learn high-response representations of major tree structures. • Precise delineation of tree boundaries through a direction-aware graph-based technique.
Keywords:
3D urban forestry
Point clouds
Tree instance segmentation
Structure awareness
Multi-task learning
AI Summary
Key information extracted from the uploaded paper, including a brief overview, abstract, background, key highlights, visual analysis, and future outlook.
Journal
U
IF:
6.7
Papers:
323
Citations:
0