Enhanced discontinuity characterization in hard rock pillars using point cloud completion and DBSCAN clustering

This paper proposes a novel strategy to characterize the discontinuities of hard rock pillars in underground mines. In this strategy, the point cloud completion technology is proposed to fill in the missing point clouds caused by defective digital photogrammetry. An improved density-based spatial clustering of applications with noise (DBSCAN) algorithm is developed for accurate discontinuity extraction. A case study is carried out to evaluate and verify the proposed strategy in characterizing discontinuities of hard rock pillars. Furthermore, the performance of the proposed method is compared with other conventional methods in extracting the pillar discontinuities, and the influence of point cloud completion on the characterization performance is discussed. The results indicate that the proposed strategy can accurately extract discontinuities, and that the proposed point cloud completion significantly improves characterization accuracy. Although the proposed strategy has not yet been applied or verified in the discontinuity extraction of other rock masses, it provides a new opinion and a reliable exploration for the discontinuity characterization of hard rock pillars.