基于摄影测量三维点云的岩体结构面自动识别方法

摘要

A method was presented to identify and analyze the plane surface of rock slope by using the 3D point cloud data obtained by digital photogrammetry The method using every point and its neighboring points constitute a subset of the plane to form fitting structure face candidate,using principal component analysis(PCA)calculation normal vector of the plane and eliminating abnormal points, recognition of discontinuities and calculate the occur-rence(dip direction and dip angle).An example of a rock slope in Beishan area of Gansu is used to identify the discontinuities.Finally,the dominant discontinuities are grouped and compared with the measured ones.The re-sults show that the recognition effect is good.%建立了一种用数字摄影测量得到的三维点云数据识别和分析岩石边坡出露的面状结构面的方法.该方法使用每一个点及其相应的邻近点构成的子集平面来形成识别结构面的候选,使用主成分分析(PCA)计算平面法向量并剔除异常点,识别结构面并计算其产状(倾向和倾角).以甘肃北山一岩石边坡为例,使用该方法进行结构面识别.最后对结构面进行优势分组;同时与实测产状进行了对比,结果表明识别效果较好.