S-Space College of Engineering/Engineering Practice School (공과대학/대학원) Dept. of Energy Systems Engineering (에너지시스템공학부) Theses (Ph.D. / Sc.D._에너지시스템공학부)
Automated Characterization of Rock Mass Discontinuities Using LiDAR Point Cloud
라이다 점군자료를 이용한 암반 불연속면 특성검출 자동화 연구
- Issue Date
- 서울대학교 대학원
- LiDAR; point cloud; rock mass classification; GSI; discontinuity characterization; automation; 라이다; 점군; 암반분류; 불연속면 특성화; 자동화
- 학위논문 (박사) -- 서울대학교 대학원 : 공과대학 에너지시스템공학부, 2020. 8. 전석원.
- The technique for determining rock mass quality and its stability is an important issue often encountered in many engineering projects including open pit and underground mines, slopes, tunnels, dams and others. Hand-mapping has been widely used as a conventional way to collect information of rock mass and determine the rock mass class. Then, a quick, safe and objective way for assessment of rock mass quality is desired to maximize the efficiency and economic benefits of the task as well as to provide essential feedback for the design, construction and operation of engineering projects. In this study, a light detection and ranging (LiDAR) technique, which can acquire 3D point cloud information quickly and accurately, was used to compensate for the shortcomings of field geological hand-mapping methods (scan line survey, window mapping survey, etc.).
The geological strength index (GSI) was assessed by quantifying the characteristics of rock discontinuity using the point cloud data obtained from LiDAR scan on rock slopes. A circular window was adopted to visually represent the distribution of rock mass quality in a target rock mass.
Prior to rock discontinuity characterization using LiDAR, the most important step is to extract the discontinuities from the point cloud. Thus, a triangulated irregular network was constructed using the ball-pivoting algorithm. Then, a patch was extracted by defining a set of triangular elements that satisfies the angle condition between adjacent triangular elements as a patch.
Patch detection performance according to the different conditions of angle and point interval was confirmed to be independently applicable to the density of different point clouds, based on the specification or measurement location and distance of the LIDAR equipment. Optimal conditions were applied for determining the orientation of the joint, smoothness, waviness, joint spacing, and block volume. The results showed a good agreement among these factors, and thus, could be applied to two sites for comparison of measurements by the LiDAR process and hand-mapping. Consequently, similar GSI values were obtained, confirming the applicability of GSI rock classification using LiDAR. After a GSI calculation employing an overlapping circular window, a technique for determining the GSI distribution was presented using the contour plot shown in the point cloud for the target.
This study aims to develop an automated algorithm that can minimize the the human bias and risk associated with field work, to quickly calculate the GSI with less manpower, and to be applied to sites requiring rapid rock engineering decisions. Another consideration is the reduction of labor and time consumed in hand-mapping. Such advantages can be maximized especially in huge survey areas or areas inaccessible targets.