Abstract: Rapid identification and stability assessment of hazardous areas on rock slopes are key issues in the safety management of open-pit mines and steep rock slopes. To address the shortcomings of traditional investigation methods, such as low efficiency, strong subjectivity, and difficulty in obtaining high-precision structural plane information, this paper proposes a method for hazardous area identification of rock slopes by integrating UAV oblique photogrammetry and automatic structural plane identification technology. Multi-view high-resolution images are acquired by UAVs to construct a three-dimensional real-scene model with centimeter-level accuracy. On this basis, octree spatial partitioning, fuzzy C-means clustering, and density-based clustering algorithms are combined to achieve automatic identification and refined classification of rock mass structural planes, and to systematically extract their occurrence, trace length, and spatial distribution characteristics. Taking a verification area of 142.3 m×30.2 m on the east slope of Yanshan as an example, a total of 78 structural planes are identified, with an average attitude of 245.4°∠45°, an average spacing of 0.78 m, and an average trace length of 1.54 m. The automatically identified structural plane parameters are further introduced into slope kinematic analysis and a three-dimensional limit equilibrium stability calculation model to comprehensively evaluate potential slope failure modes and their safety under different working conditions. The results show that the dip directions and dip angles of structural planes in Yanshan approximately follow normal distributions, with coefficients of variation generally lower than 20%, indicating stable development of the major structural plane sets. In contrast, the trace lengths of structural planes exhibit a lognormal distribution, with coefficients of variation exceeding 100%, reflecting significant rock mass heterogeneity and pronounced differences in joint development scales. Stability analysis indicates that potential unstable blocks are concentrated within the study area and are mainly structure-controlled blocks with heights smaller than a single bench height, thicknesses less than 3 m, and exposed areas less than 20 m². Under natural conditions, the proportion of blocks with low safety factors is generally less than 10% of the total identified blocks; under water-bearing or external disturbance conditions, the overall safety factors of blocks decrease by approximately 13.6%-15.5%, and the proportion of unstable blocks increases significantly, indicating that water action and engineering disturbance are important controlling factors inducing rock slope instability. Therefore, the proposed method can effectively reveal, in a statistical sense, the intrinsic relationship between rock mass structural characteristics and the distribution of hazardous slope areas, enabling rapid identification and quantitative assessment of hazardous areas on rock slopes, and providing reliable technical support for safety monitoring and risk zoning of open-pit mines and similar rock slopes.