利用偏度平衡自动提取机载LiDAR点云城区道路

针对机载LiDAR点云提取城区道路自动化程度低以及提取道路网不完整的问题,提出一种基于偏度平衡算法自动提取城区道路的方法。首先,计算末次回波点云中邻近点的坡度,依据城市道路设计标准选择坡度阈值,分离非地面点并获得初始道路点云;然后,利用偏度平衡算法计算出最优强度阈值,滤除非道路点,再结合道路与停车场的空间位置关系,通过点距精化初始道路点云;最后,对道路区域进行细化和平滑,得到道路中心线。通过定量分析及对比实验,该方法能够自动地从LiDAR点云中提取出较为完整的道路网。

Urban Road Extraction from Airborne LiDAR Point Cloud Automatically Based on Skewness Balance Algorithm

Aiming at the problem of low degree of automation and incomplete extraction of urban road from LiDAR point cloud,a method of extracting road automatically based on skewness balance algorithm is presented.Firstly,unground points are removed and the initial road point cloud is obtained through the comparison of the slope between the adjacent points calculated in the last echo point cloud with the slope threshold according to city road design criteria.Secondly,optimal intensity threshold is calculated using skewness balance algorithm to filter out nonroad points,and the initial road points are refined by the points distance.Finally,the centerline of the road is extracted though thinning and smoothing.The experimental results show that the proposed method can extract a more complete road network from the LiDAR point cloud automatically.