TreeScope is the first semantically segmented LiDAR dataset collected with robotic systems in agricultural environments.

Overview

TreeScope provides LiDAR data and ground-truth data for semantic segmentation and diameter estimation from agricultural environments to address the counting and mapping of trees in forestry and orchards. Raw data is available from UAV flights and our custom sensor platform, including sensor data from LiDAR, IMU, GPS, RGBD and thermal cameras. We provide ground-truth in the form of manually annotated semantic labels for the tree stems and field measurements of tree diameters.

TreeScope processed data, raw data, ground-truth data, benchmark scripts, and code are available to download. Check out our Github repo for an overview on how to use the data and benchmark scripts to evaluate the performance of diameter estimation and semantic segmentation algorithms.

License

TreeScope is distributed under the Creative Commons Attribution-NonCommercial-ShareAlike 4.0 International License. This means that you can copy and redistribute this dataset, as well as remix, transform and build upon this work. If you use TreeScope, please cite our paper. You cannot use this dataset for commercial purposes. Finally, any modifications should be released under the same license.

Paper

You can access the paper from arXiv. To cite our work, please use:

@misc{cheng2023treescope,
  title={TreeScope: An Agricultural Robotics Dataset for LiDAR-Based Mapping of Trees in Forests and Orchards}, 
  author={Derek Cheng and Fernando Cladera Ojeda and Ankit Prabhu and Xu Liu and Alan Zhu and Patrick Corey Green and Reza Ehsani and Pratik Chaudhari and Vijay Kumar},
  year={2023},
  eprint={2310.02162},
  archivePrefix={arXiv},
  primaryClass={cs.RO}
}

Acknowledgements

We gratefully acknowledge the support of the IoT4Ag ERC funded by the National Science Foundation (NSF) under NSF Cooperative Agreement Number EEC-1941529, NIFA grant 2022-67021-36856, and NSF grant CCR-2112665.