关键词: 高位危岩体, 机载LiDAR, 无人机影像, 稳定性, 成因分析

Abstract: In water conservancy and hydropower projects, dangerous rock masses often develop in hidden areas of high and steep slopes, posing challenges for accurate surveying and measurement using conventional methods. Swiftly and conveniently identifying dangerous rock masses on such slopes is of significant engineering importance. To address this issue, we take the investigation findings of dangerous rock masses outside the opening line of slope for critical building at Huangdeng Power Station as the research basis. We constructed a 1∶500 scale 3D model of the actual scene using UAV (unmanned aerial vehicle) oblique photography. By using airborne LiDAR remote sensing measurement technology, we acquired high-precision LiDAR point cloud data and optical image data in the research area. By incorporating a 3D visual remote sensing interpretation platform, we established the interpretation marks for dangerous rock mass, allowing for the extraction of structural plane characteristics of these rock masses. Furthermore, we conducted in-depth field investigations and review of the interpretation results. Our review work confirms the robust accuracy of remote sensing measurement technology in the initial stage, with highly credible interpretation results. Furthermore, by employing a geospatial data analysis method, we explored the development distribution patterns and genesis mechanisms of dangerous rock masses. The findings demonstrate that dangerous rock masses in the study area predominantly develop in the elevation range from 1 609 m to 1 963 m and the slope gradient range from 40° to 70°, primarily facing the south direction. Additionally, dangerous rock disasters is also affected to some extent by river systems and human engineering activities. The research findings hold significant reference value for the identification and causal analysis of high-risk rock masses in hydropower reservoir areas.

Key words: dangerous rock mass, airborne LiDAR, unmanned aerial vehicle, stability, cause analysis