A method for identification rock mass discontinuities in underground drift with pre-separation of linear and planar point cloud features
Zhongyuan Gu, Xin Hong Xiong, Chengye Yang, Miaocong Cao · Ain Shams Engineering Journal · 2024
• Through the point cloud filling and voxel filtering down-sample, the point cloud of rock mass is obtained with better quality and more concise. • A pre-segmented discontinuities identification method P-RG is proposed, which has better accuracy compared with the traditional RG algorithm. • This paper presents a method to identify the discontinuities based on point cloud of rock mass in underground drift. Feature extraction is a common method for obtaining the discontinuities of rock masses from point cloud. In this study, a feature pre-segmentation method for identifying the discontinuities from point cloud obtained from underground drift is proposed. Specifically, the proposed method involves: (1) acquiring underground drift point cloud using 3D laser scanning; (2) obtaining a more complete, regular, and simplified drift point cloud by filling void holes and voxel filtering down-sample; (3) proposing a P-RG algorithm to pre-segment linear and planar features of drift point cloud, which enables better identification of discontinuities and ultimately obtaining the discontinuities and the orientation from the point cloud; and (4) using the DBSCAN clustering algorithm to cluster the dominant group. The proposed method was tested using the underground drift of a gold mine in Jilin Province, China, and compared with traditional feature extraction methods, demonstrating its accuracy and efficiency in identifying discontinuities in underground drift.