中文摘要：

机载LiDAR平台扫描得到的地形数据点具有高密集度,简化处理大规模地形数据而又不牺牲数据精度的特点而成为研究重点。利用地表流水物理模拟方法提取粗略的矢量地形特征线,再对特征线做缓冲区分析,对缓冲带中包含的LiDAR点云数据构建三角网,优化提取三角网的有效结点,以确定区域的地形特征点。由此得出的地形特征点既简化了LiDAR点云数据,又清晰地描述了地形结构。

英文摘要：

The topographical data points scanned from airborne LiDAR platform have high density, so simplifying numerous topographic data, yet without sacrificing data accuracy becomes a research focus. At first, extracting rough terrain feature lines by surface water physics simulation, then the feature lines are used to do the buffer analysis, and LiDAR points in the buffer zone are used to construct the triangulation and take the re-extract nodes to gain terrain feature points. The result not only simplifies the LiDAR point cloud data, and clearly describes the terrain structure.