机载LiDAR点云冗余数据辨识及消除

he repeated measurement in strip overlap areas and the duplicate records in multiple echoed are the error source for multiple-return LiDAR point cloud redundant data. The existence of both will result in increased data volume, reduce the efficiency of data processing, and the latter will also result in uneven distribution of point cloud. For the former, a method for redundancy elimination based on neighboring-point is proposed. The proposed method defines, find the neighboring-point, which consists of two points from different strips and whose point spacing is smaller than a given threshold. And then the low accuracy laser point is removed to eliminate redundancy. For the latter, a method for redundancy elimination based on corresponding points is proposed. The proposed method compares the distance between each two echoes of single laser pulse, if the distance equal to 0 then the relatively later echo is removed to eliminate redundancy. The feasibility and effectiveness of the proposed method is verified using AnYang and ISPRS test data.