Forestry

Detect canopies and tree trunks with the forestry AI model

The staring point is a quality point cloud classification (semantic segmentation). For this task, Flai trained a proprietary AI model specifically for forestry applications. This model is available directly in Flai web application and can be simply applied to imported point cloud.


With machine learning on aerial LiDAR point clouds, the following classes are added to the forestry AI model:

  • Ground

  • Low understory

  • Medium understory

  • High understory

  • High vegetation and canopy

  • Tree trunk

  • Fallen trees


With a classified point cloud, we are able to deliver additional outputs that are useful for forest inventory and carbon trading:

  • Tree top detection,

  • Trunk detection with fitted 3D lines.

  • Segmentation of canopies with tree height and canopy area.

Industry gallery