利用地形开度算子自动提取激光雷达点云断裂线

Automatic Break Line Extraction from Airborne LiDAR Point Cloud Using Topographic Openness

基于机载激光雷达(Li DAR)点云生产高质量的数字高程模型(DEM),需要地形特征线的约束。本文针对现有断裂线提取算法仅仅考虑了地表的局部变化,错提取率高的现状,从人工解译的角度出发,引入地形开度(topographic openness)作为一种定量描述地表整体变化的地形特征,提出了基于地形开度的断裂线自动提取算法。首先将原始地面点按高程采样成距离图像,然后通过地形开度算子获取地表的主体结构变化,最后结合形态学算子和边缘提取算子提取断裂线种子点并跟踪断裂线矢量。试验结果表明,该算法无需人工干预,能够快速提取较完整的断裂线。

In order to produce high quality digital elevation model（ DEM） based on airborne lidar（ LiDAR） point cloud,break lines need to be used as constraints.Aiming at the fact that error rate of break line extraction is relatively high because of existing algorithms only consider the local variation of the earth＇s surface,topographic openness,starting from the manual interpretation point of view,is applied as a new operator to extract breaklines. Firstly,the original ground points are sampled into the distance images. And then the main structure changes of the earth surface are extracted by topographic openness. Finally,the seed points of the break lines are extracted by morphological operators and edge extraction operators,and the vector of the broken lines is tracked.The experimental results show that the algorithm can quickly extract complete break lines without manual intervention.