Automatic extraction of rock discontinuities from point clouds using dynamic DBSCAN algorithm

Abstract

Detection and mapping of rock discontinuities are important during excavation. For the acquisition of precise quantitative data, terrestrial laser scanning (TLS) technology is widely used. However, studies on the effect of discontinuities parameters on detection are limited. We created discontinuity models with various roughnesses and connectors with various angles using 3D printing technology. Therefore, we can control the variables in the scanning. Several open-source packages were employed to obtain the information from the point cloud acquired by TLS. The result shows that as the angle between discontinuities increases, the recognition effect decreases. Furthermore, the roughness of the discontinuity makes it more likely to cause incorrect classification during the detection process. The effectiveness and limitations of structural plane information algorithms based on TLS technology to guide the engineering application of the technology are discussed in this study.