Abstract
Data from the article "Point Cloud Data Mining with HD Map Priors for making Synthetic Forest Datasets" by K. Karlauskas, J. Gelšvartas, P. Treigys (2025) Article: https://doi.org/10.1109/JSTARS.2025.3593827 Code repository: https://github.com/kasparas-k/pointcloud-tree-data-mining # Dataset contents 1. individual_trees.zip -- individual tree candidate point clouds, including erroneous or otherwise noisy segmentation results. The file `filtered_features.json` contains geometric features of all automatically extracted clusters, while `checked_features.json` contains geometric features only of a subset of visually vetted tree instances. 2. synthetic_scenes.zip -- synthetic forest scene *.laz point clouds with individual tree instance labels in the scalar field treeID 3. segmentation_test_output.zip -- output segmentation labels with various algorithms, refer to the original publication. 4. Data of individual tree locations in Tallinn and Warsaw acquired from https://www.openstreetmap.org/ is attached in the `openstreetmap` folder. The geolocation data from OpenStreetMap is licensed under ODBL https://opendatacommons.org/licenses/odbl/