边缘系数建筑物雷达点云数据简化方法

A data compression method of LiDAR point clouds data with boundary coefficients

针对基于LiDAR点云数据进行建筑物自动重建中存在的数据冗余问题,该文设计了一种定量描述激光点位于地物边缘区几率大小的指标——边缘系数,并据此提出了基于边缘系数的建筑物LiDAR点云数据简化方法。该方法利用激光点与其邻域点的位置、数量及分布计算该点的边缘系数,通过试验分析确定边缘系数的阈值并对点云数据进行分割,最后保留建筑物边缘区域的点,实现点云数据的简化。实验表明,该方法在对点云数据进行高效压缩的同时有效保留了位于地物边缘处的点云,有助于提高海量点云数据处理能力和建筑物重建效率。

There is much more redundancy in LiDAR point cloud data when it has been used to reconstruct buildings.So the reasonable compression of LiDAR point cloud data will improve the efficiency and accuracy of building reconstruction.In this paper a boundary coefficient was proposed,which could estimate the possibility of a point located near the building boundary.Based on the boundary coefficient a new LiDAR data compression method was proposed.The border coefficient was calculated out of consideration for point position relative to other points and the amount distribution of the points in a nearest region.And a boundary coefficient threshold evaluated by experiment was used to segment the point clouds,then the point clouds compression which keeps the border point persisted was realized.Experimental result showed that this method could compress the point clouds effectively and persist the building boundary points at the same time,and will conduce to massive point clouds data processing and building reconstruction.