Identification and automatic recognition of discontinuities from 3D point clouds of rock mass exposure

Mapping and analyzing rock mass discontinuities based on 3D (three-dimensional) point cloud (3DPC) is one of the most important work in the engineering geomechanical survey. To efficiently analyze the distribution of discontinuities, a self-developed code termed as the cloud-group-cluster (CGC) method based on MATLAB for mapping and detecting discontinuities based on the 3DPC was introduced. The identification and optimization of discontinuity groups were performed using three key parameters, i.e. K , θ , and f . A sensitivity analysis approach for identifying the optimal key parameters was introduced. The results show that the comprehensive analysis of the main discontinuity groups, mean orientations, and densities could be achieved automatically. The accuracy of the CGC method was validated using tetrahedral and hexahedral models. The 3D point cloud data were divided into three levels (point cloud, group, and cluster) for analysis, and this three-level distribution recognition was applied to natural rock surfaces. The densities and spacing information of the principal discontinuities were automatically detected using the CGC method. Five engineering case studies were conducted to validate the CGC method, showing the applicability in detecting rock discontinuities based on 3DPC model.