一种基于平面拟合的LIDAR点云滤波方法

LIDAR点云滤波是将LIDAR点云数据中的地面点和非地面点分离的过程。根据在较小区域内可以近似认为地面为一平面,本文提出了一种应用平面拟合的方法,首先在一个局部区域内拟合出一个近似平面,通过判断LIDAR点是否属于该平面来获取平面点,并通过分类处理从平面点中得到地面点,最后用得到的地面点内插出DEM。滤波前,需要剔除高程异常点,本文应用了高程差约束算法抑制高程异常点,从而较好地保持了原始数据的局部细节信息。

A method for filtering LIDAR points cloud based on planar fitting

LIDAR points cloud filtering is a preprocessing for the application of LIDAR data,it can filter out non-ground data from original data set.On the assumption that ground plane can be considered a planar plane in a small area,this paper developed a method that fits a approximating plane first using points in the area,and then acquires point on the plane by judging whether a point belonging to the plane,finally,using points on the ground to generate a DEM.The spikes and pits caused by anomalies should be remov...