Identification of Unstable Rock Masses Based on 3D Point Cloud Processing

高位危岩体因高差显著、坡度陡峭，难以通过人工直接接触调查。文章以桂林市临桂区会仙镇大神山石灰岩矿山边坡危岩体为例，采用无人机摄影测量的方法进行数据采集，通过影像后期处理软件对点云数据处理后，采用DBSCAN点云聚类算法，将点云区分为危岩体点云和非危岩体点云，实现对危岩体的识别与范围界定。研究成果对高位危岩体的调查与评价具有一定参考价值。High-position unstable rock masses are challenging to investigate through direct manual surveys due to their significant elevation differences and steep slopes. Taking the limestone quarry slope at Dashan Mountain in Huixian Town, Lingui District, Guilin City, as an example, this study utilized UAV photogrammetry for data acquisition. After processing the point cloud data using image post-processing software, the DBSCAN clustering algorithm was employed to classify the point clouds into unstable rock mass points and non-unstable rock mass points, thereby achieving the identification and boundary delineation of unstable rock masses. The research findings provide valuable references for the investigation and evaluation of high-position unstable rock masses.