samstars Documentation¶
Multimodal individual tree crown segmentation from imagery and lidar-derived rasters.

Overview¶
samstars is a Python package for instance segmentation of individual tree crowns from high-resolution imagery and lidar-derived rasters. The resulting crown polygons can be used for downstream species classification.
It combines:
- high-resolution imagery (satellite, aerial, or drone)
- lidar-derived raster layers (e.g., canopy height and density)
It automatically derives prompts from lidar structure and spectral cues to segment individual tree crowns with limited training annotations.
First Segmentation¶
Use the model bundle created by Training with imagery and lidar rasters prepared as described in Data Preparation.
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 | |
See Installation to install the library. For a minimal training and segmentation workflow, see Quick Start. For task-focused documentation, see Data Preparation, Training, Segmentation, and Downstream Analysis.