基于曲率统计的LiDAR点云二次滤波方法

针对传统偏度平衡方法滤波结果中存在低矮植被、建筑物侧面墙脚等非地面点云问题,在传统偏度平衡方法的点云一次滤波算法的基础上,提出一种基于曲率统计的点云二次滤波方法。对该方法进行试验,并将试验结果与传统偏度平衡方法滤波结果进行对比分析。结果表明:基于曲率统计的点云二次滤波方法比传统偏度平衡法能够多滤除83%的植被点云、5%的建筑物点云,能够有效地滤除传统偏度平衡方法滤波结果中的低矮植被、建筑物侧面墙脚等非地面点云。

A secondary filter method of LiDAR point cloud based on curvature statistics

The rusults of conventional skewness balancing exist non-ground points of low vegetation and side of buildings. Aimed at this, a secondary filter method of LiDAR point cloud was developed based on curvature statistics to filter the remained non-ground points in the result of filter based on conventional skewness balancing. The results of the secondary filter method proposed and conventional skewness balancing were compared. The results show that the former can remove 83% more pionts of vegetation and 5% more pionts of building than the latter. The secondary filter method of LiDAR point cloud based on curvature statistics can efficiently filter the non-ground points remained in the conventional skewness balancing method.