A Filtering Algorithm of Airborne LiDAR Points Cloud Based on Least Square

摘要

Filtering algorithm based on least square method can get digital elevation model of high precision, but the filtering effect is poor if non-ground point set is big. In order to solve this problem, points of smaller elevation are used to initial curved surface fitting, which has smaller calculation compared with fitting by all points. The curved surface is closer to the ground, making it conducive to the subsequent iteration. Only ground point set is processed in the subsequent iteration, so the processing is simple. The experimental results show that the algorithm has smaller error and can filter out large part of non-ground points effectively.