A High-Resolution UAV Color and CHM Dataset for Bamboo Forest Segmentation and Mapping

The Bamboo Remote Sensing Dataset comprises co-registered UAV RGB imagery, LiDAR-derived canopy height models (CHMs), and pixel-wise annotated masks for bamboo forest mapping. It is designed to support research in semantic segmentation, multimodal data fusion, and UAV-based forest monitoring, and is suitable for both classical and deep learning–based image analysis. The dataset enables applications including bamboo mapping in unmanaged forests, multimodal fusion studies, segmentation benchmarking, and ecological monitoring. To facilitate reuse and reproducibility, trained UNet models for bamboo segmentation using single-modality (RGB or CHM) and fused multimodal inputs are also provided. Additional details on data organization, preprocessing, and model usage are available in the accompanying \texttt{README.md}. Copyright (c) 2026 Chinsu Lin Permission is granted for: ✓ Academic research✓ Educational use✓ Non-commercial personal use Commercial use, redistribution, sublicensing,or incorporation into commercial productsrequires prior written permission from the copyright holder. For commercial licensing, contact: Email: [e-mail]