.png){kind=icon}
.png){kind=icon}
Return
A synthetic data generation framework for deep learning-based LiDAR forest structure analysis
.png){kind=icon}
PRE
AI
J
D
H
C
J
D
DOI：10.1016/j.rse.2026.115436.png)
.png){kind=icon}
Original request for help.png){kind=icon}
Share.png){kind=icon}
Save
Abstract
En 中文
• Automated framework generates diverse multi-platform forest LiDAR data. • Boreal3D: A fully annotated synthetic forest-plot dataset for four platforms. • Synthetic pre-training enables efficient and transferable 3D forest analysis. • Fine-tuning with only 20% real data matches full real-data training performance. • Provides a scalable benchmark for advancing large-scale forest 3D structure studies.
Keywords:
synthetic data
LiDAR forest structure
deep learning
multi-platform
benchmarking
Journal
IF:
11.4
Papers:
1.1W
Citations:
9.4W
Organization
Cited Papers