Abstract: In order to accurately identify high and steep slope dangerous rock masses in karst mountainous areas, the authors took the typical karst landform area of Luocheng in Guangxi as an example, through three low-altitude unmanned aerial vehicle remote sensing techniques, that is airborne LiDAR, oblique photogrammetry, and close-range photogrammetry, to conduct detailed investigation of dangerous rock masses. The extracting information methods for isolated rocks and cracks under forests were summarized based on point cloud data, and the 3D real scene models were systematically summarized for extracting structural plane attitude information and structural plane characteristic parameter information. The identification and refined investigation methods of dangerous rock masses on high and steep slopes were discussed, and the refinement identification of dangerous rock masses was carried out, with a total of 34 dangerous rock bodies being identified. The research results could provide technical guidance for identifying hazardous rock masses in similar regions.