Research on rock slopes has attracted considerable attention in the fields of infrastructure construction and disaster prevention. The search for an efficient and reliable method for discontinuity collection is an important task for various studies of jointed rock slope. Capturing high-precision and complete rock exposure images based on unmanned aerial vehicle (UAV) multi-angle and nap-of-the-object photogrammetry technology is proposed to realize the construction of a millimeter-level 3D point cloud for the high and steep slope. This study develops a new automatic methodology for discontinuity identification and interpretation. A Delaunay triangulation network is created for connecting adjacent points in the point cloud to calculate normal vectors and search for neighbor points efficiently. Combined with the modified region-growing algorithm, an automatic procedure which identifies discontinuities and acquires geometrical parameters efficiently and accurately is established. The proposed approach is applied to the high-steep rock slope on the northern bank of the Sequ Grand Bridge in the Changdu area of Tibet. Compared with the results from the manual survey, the proposed method is more reliable and exhibits high accuracy for discontinuity identification and interpretation on rock exposure.
Read full abstract