A Method for Extracting Street Trees from Mobile LiDAR Point Clouds

Abstract

This paper presents a method to extract street trees from laser scanning point clouds based on segmentation and growth of three dimensional (3d) grids. We firstly build the 3d virtual grid of scanning area point clouds. Then, the distribution characteristic of point cloud is analyzed in a grid cell (density, elevation etc.). Finally, the point clouds of tree trunk and tree crown are extracted through grid segmentation and growth by a bottom-to-top way. In order to verify the validity of this method, we used the designed method to analysis the real collection of LiDAR point cloud data.