Preliminary Identification And Evaluation Of Dangerous Rock Masses On High And Steep Slopes Based On UAV Photography

Conventional,human-based methods of surveying dangerous rock masses are very risky and time consuming due to problems regarding traffic inconvenience,steep terrains and unstable rock masses.Because of the advantages of flexibility,safety and low cost,the use of unmanned aerial vehicles(UAVs)to survey dangerous rock masses is gradually replacing manual work.This paper describes a method for the preliminary identification and stability assessment of dangerous rock masses based on UAV photography technology,and assesses its applicability using the point cloud model of the bank slopes of the Lianghekou Hydropower Project.The point cloud model can accurately express the characteristics of the dangerous rock mass on the slope.UAVs are used to perform multi-angle overlapping photography for the study area.The motion recovery structure algorithm(SFM)is used to process aerial photos to generate a dense point cloud model.The characteristics of dangerous rock masses are parameterized to calculate these characteristics using point clouds.Considering the characteristics of the boundary points of dangerous rock masses,the feature points of the model are estimated based on the discrete curvature.The density-based clustering algorithm(DBSCAN)is used to identify different point clusters,and the support vector machine(SVM)is used to extract the boundary points of the dangerous rock masses.Considering the overall regional characteristics of the dangerous rock mass,nondangerous rock mass surface points in the vicinity of the boundary point cluster of the dangerous rock mass are fitted with a plane to cut the point cloud model.After the intersection of the plane and the point cloud model is deleted,using DBSCAN algorithm to identify clusters of isolated points in the remaining point cloud.The characteristics of each point cluster are calculated and the point clusters representing dangerous rock masses are identified.The preliminary evaluation methods for the stability of dangerous rock mass regions were discussed.The stress reduction tensor theory was used to estimate the distribution of the shear stress and normal stress ratio caused by the dead weight of the dangerous rock mass on the bottom slip surface,and the rock failure location on the bottom slip surface was determined according to the Mohr-Coulomb failure criterion.The ratio of the total area,that is,the coefficient of instability N,can divide the stability of dangerous rock masses into 3 levels according to its value.The final result is given in the form of a stability rating map of the dangerous rock masses.The results show that,using UAV photography,effective identification and preliminary stability assessment of dangerous rock masses can be carried out.