Abstract:

Aiming at the problem of single tree extraction in forestry inventory,a method based on Airborne LiDAR point cloud data is proposed,which uses local maximum value and noise elimination to obtain crown vertex,and uses watershed segmentation method to extract single tree,estimating DBH and forest volume based on tree height and forest volume model.Compared with the traditional artificial field survey method,it has higher time efficiency and reduces a lot of labor cost.

Key words: airborne LiDAR, local maximum, noise elimination, watershed segmentation, single tree detection