Trees have a complex geometrical structure that is hard to capture using traditional means. Deep learning models are highly capable of learning these complex semantics and can produce superior results. Classifying trees from point cloud data is useful in applications such as high-quality 3D basemap creation, urban planning, and forestry workflows.
To complete this workflow, the following are the license requirements:
- ArcGIS Desktop—ArcGIS 3D Analyst extension for ArcGIS Pro
This model has the following characteristics:
- Input—Point cloud in LAS dataset file format (.lasd) or as per the Classify Point Cloud Using Trained Model tool.
- Output—Classified point cloud with two classes: tree/high-vegetation and background.
- Compute—This workflow is compute intensive and a GPU with compute capability of 6.0 or higher is recommended.
- Architecture—This model uses the PointCNN model architecture implemented in ArcGIS API for Python.
- Extra attributes—Input point clouds should have at least X, Y, Z, and Number of Returns.
- Accuracy metrics—For the validation dataset, the F1-score is 0.970628.
- Class mapping—Points predicted as tree/high-vegetation are mapped to class code 5, and those predicted as background are mapped to class code 0.
Access and download the model
Download the Tree Point Classification pretrained model from ArcGIS Living Atlas of the World. Alternatively, access the model directly from ArcGIS Pro.
The following are the release notes: