基于DEM匹配的滑坡检测

Landslide Detection Based on DEM Matching

基于不同时期DEM数据检测地表变形是进行滑坡、泥石流等地质灾害检测的有效手段,DEM匹配是关键,两个变形地表的模型曲面特征对应关系确立是难点。针对高山地区DEM匹配问题,提出了一种新颖且稳健的基于等高线和最小二乘曲面匹配相结合的方法。该方法首先利用等高线的形状检测地形没有变化的山顶区域,提取位置不变的山头点作为控制点,能够提供足够好的转换参数初始值;其次采用最小二乘曲面匹配技术实现最优匹配。试验结果表明所提方法不受DEM粗差、数据异常以及地形变形的影响,具有很强的抗噪特性,可以在复杂地形条件下检测出稳定不变的控制点,取得较好地配准效果,最终在高程差异图上检测出明显的滑坡区域。

Terrain deformation detection based on DEM data from different periods is an effective method to detect landslides, debris flows and other geological disasters. DEM matching is the key and it is difficult to establish the corresponding relationship between the surface features of the two deformed terrains. To solve the problem of DEM matching in variable mountainous areas, a new and robust method based on contour-based point matching and least squares surface matching is proposed in the paper. Firstly the shape of contour lines can be used to detect the area without terrain change and invariant terrain peaks can be detected as control points, which can generate good initial values of the transformation parameters. Secondly, the least squares surface matching technique is used for optimal matching. The experimental results show that the proposed method is not affected by DEM gross error, data outliers and terrain deformation, and has strong robustness to noise. The stable and invariant control points can be detetced in complex terrains, and the good registration result is archieved. Finally, the obvious large-scale landslide areas can be detected in the elevation difference map.