鞍部点是重要的地形控制点之一,由于其特殊的形态特征,基于DEM的鞍部点提取算法较山顶点、山谷点提取更为困难。本文提出一种基于DEM汇流模拟的鞍部点快速提取改进算法。该算法利用DEM地表汇流模拟的方法提取山脊线,利用地形倒置得到的反地形地表汇流模拟提取反地形流域边界线,获得能自动实现地形剖分的原始地形汇水线,进而通过特征线求交实现鞍部点的快速自动提取。不同地貌样区实验显示,该方法摆脱了传统移动窗口分析方法无法顾及地形整体变化的缺点,考虑了鞍部点与周围地形特征点线之间的拓扑关系,提高了鞍部特征点的提取精度。
The saddle is one of the most important terrain feature points, the extraction of saddles based on DEMs is of significance to digital terrain analysis. Nevertheless, owning to saddles'unique morphological characteristic, it has more difficulties in extracting saddles than any other terrain feature points. This paper presented an improved saddles extraction method based on runoff concentration simulation in DEM. Firstly, the ridge lines were extracted by runoff concentration simulation in DEM. Then, inversed DEM were derived through overturning original landform to extract catchment lines, which can automatically divide the terrain. Finally, the saddles were extracted from intersecting feature lines. The experiments in different study areas demonstrated that this improved method considered the topological relationship between the saddles and surrounding terrain features, overcoming the disadvantage that the neighborhood analysis cannot take consideration of the whole terrain change. And the accuracy of saddles extraction was improved.