基于LiDAR点云数据的水体轮廓线提取方法研究

A Method for Extracting Water Contour Lines from LiDAR Point Clouds Data

提出一种基于机载激光雷达点云数据提取水体轮廓线的方法。采用双层格网模式提取较窄的水体;以朝向水体的边界点作为拟合轮廓线的关键点提取更精确的轮廓线。实验表明,该方法可以较好地提取水体轮廓线。

A contour lines extraction method for 3D water body from LiDAR point clouds data is proposed in this paper. First, all points are divided into rough and equidistant grids whose number, average elevation and height standard deviation of every grid points will be calculated later. Then, confirm the grids belonged to specific water areas with strict conditions according to four characters of the point clouds in the waters： the low density, the similarity of elevation in all points, the weak echo intensity, water surface is lower than the surrounding land. Next, exclude sparse and discrete blocks through the connected domain by the character of larger water areas. Finally, re-organize point clouds data by fine grid. Extract water areas with the seed filling method under more relaxed conditions starting from the grids belonged to the specific water areas. Experiments indicate that this method is feasible.