特征区域最近点迭代地表移动分析算法

Surface Movement Analysis Algorithm Based on Feature Region Closest Points Iteration

为了获取人员难以到达的陡坡、悬崖等危险地形的地表移动和变形值,采用三维激光扫描技术对其进行精细测量,结合点云特征区域提取和最近点迭代(Iteration Closest Points,ICP)算法,对点云分析方法进行了深入研究,提出了特征区域最近点迭代地表移动分析算法(Surface Movement Analysis Algorithm Based on Feature Region Iteration Closest Points,FRICP)。该算法通过点云滤波处理,实现地面点和植被点有效分离;在此基础上,将法向量和植被高程信息相结合,提取点云中的特征点;再对同一区域地面特征点进行聚类分析,形成特征区域;然后建立球形搜索区域及特征度量指标,匹配多期同名特征区域;最后使用ICP算法对同一地点不同时期的观测点云进行计算,利用获得的坐标变换参数求取地表移动值。采用某矿开采沉陷区山坡移动监测数据进行了算法验证,结果表明:FRICP算法可以准确获取山地边坡的移动和变形值,与人工精确判读结果相比,三维坐标最大偏差为12 mm,最小偏差为3 mm。FRICP算法可同时计算地表下沉和水平移动值,为利用三维激光扫描技术分析山地边坡地表移动提供了新思路,可为山区开采沉陷监测等领域提供技术支持。

In order to obtain the surface movement and deformation values of dangerous terrains such as steep slopes and cliffs,which are difficult for personnel to reach,three-dimensional laser scanning technology is used for fine measurement. The point cloud analysis method is further studied by combining the extraction of point cloud feature region and iteration closest points(ICP) algorithm. A surface movement analysis algorithm based on feature region iteration closest points(ICP) is proposed. The algorithm realizes the effective separation of ground points and vegetation points by point cloud filtering. On this basis,the normal vector and vegetation elevation information are combined to extract the feature points in the point cloud. Then,cluster analysis is carried out on the ground feature points in the same area to form the feature region. Thirdly,the spherical search area and feature measure index are established to match the multi-phase feature region with the same area. Finally,the ICP algorithm is used to calculate the observation point cloud at the same location in different periods,and the surface movement value is obtained by using the calculated coordinate transformation parameters. The FRICP algorithm is verified by monitoring data of slope movement in a mining subsidence area. The results show that the FRICP algorithm can accurately obtain the mountain ground movement. Compared with the manual accurate interpretation results,the maximum deviation of 3D coordinates is 12 mm and the minimum deviation is 3 mm. The FRICP algorithm can calculate the surface subsidence and horizontal movement at the same time,which provides a new idea for analyzing the surface movement of mountain slope with three-dimensional laser scanning technology,and can provide technical support for mining subsidence monitoring in mountainous areas.