Abstract: Abstract:[Objective]There are a large number of high and steep slopes in mountainous areas in my country. Due to their hidden and dangerous characteristics, it is currently difficult to accurately obtain the distribution location and characteristic information of rock masses on high and steep slopes through manual surveys and single non-contact measurements. [Methods]This paper fuses point cloud data obtained from airborne LiDAR, ground LiDAR and UAV oblique photogrammetry with multi-source data to complement each other's advantages. The fused point cloud is then used to analyze high and steep slope hazards. Extract information from the scale boundary, trailing edge characteristics, occurrence information and structural plane characteristic parameters of the rock. [Results]The results show that the multi-source data fusion method used in this article effectively complements the advantages of various data. The fused point cloud is used to extract scale boundary, trailing edge information and structural plane characteristic parameter information of dangerous rocks on high and steep slopes. The extraction accuracy meets [Conclusion]The relevant specification requirements provide a technical method for the refined extraction of geometric parameter information of rock mass on high and steep slopes in mountainous areas.