Multi-temporal high-density terrestrial laser scanning (TLS) datasets are processed to delineating possible movements from terrain surfaces and trees. Terrain surface movements are estimated with the help of segmentation and random sample consensus (RANSAC) algorithm. Tree movements are interpreted by iterative closest point (ICP) solved translations and rotations of tree point clouds. The capabilities of the proposed methodology were tested using a case study of the Slumgullion landslide, where the trees without clear trunks cover the terrain surfaces. The displacements from the terrain surfaces and trees are similar with the results observed using our global positioning system (GPS) and historic results.

处理多时相高密度地面激光扫描（TLS）数据集以描绘出来自地形表面和树木的可能移动。借助分段和随机样本一致性（RANSAC）算法估算地形表面运动。通过迭代最近点（ICP）解决的树点云平移和旋转来解释树的运动。使用Slumgullion滑坡的案例研究测试了所提出方法的功能，其中没有清晰树干的树木覆盖了地形表面。地形表面和树木的位移与使用我们的全球定位系统（GPS）观察到的结果和历史结果相似。