Abstract: In karst mountainous area of Guizhou, in China, traditional ground survey methods often have limitations due to the existence of complex mountains with large elevation differences. In order to effectively identify and measure potential high-order position and hidden geological hazards, the technology method of integrating unmanned aerial vehicle LiDAR with ground 3D laser scanning is adopted. Through the fusion of “vertical view” data and “front view” data, the high-precision point cloud and 3D model data of the collapse can be obtained completely and accurately. This study takes the collapse of Dujiazhai in Jichang Town, Shuicheng District, Guizhou Province as an example, and finally obtains high-precision 3D model data in the study area through the steps of field data collection, original point cloud preprocessing, different data fusion, and overall coloring repair. Based on the 3d model data, the rock mass structure and fractures were extracted, and the stratification plane and two groups of joint fractures were interpreted, in order to effectively identify the spatial distribution of the dangerous rock mass. The size of the dangerous rock belt was calculated to be about 66,000 m³ based on the grid division method. The results show that the technical method of the integration of UAV-borne LiDAR and ground 3D laser scanning can complement each other with advantages, and has the characteristics of strong operability, high accuracy and accurate identification. It can effectively obtain the distribution and scale of dangerous rock mass, provide a basis for subsequent stability analysis and risk assessment.