Identification And Analysis Of Structural Surfaces Of Cis-Layered Rocky Slopes Based On Improved Area Growth Method

The spatial location and development status of these structural surfaces affect the form of deformation and damage of the rock mass and the stability of the slope,so it is crucial to collect statistical structural surface information quickly and accurately.Traditional survey means such as compass measurement method and scanning line method are low in efficiency and high in intensity,which can hardly meet the demand of rapid survey and construction of large slope projects.Close-up photogrammetry and 3D laser scanning technology are gradually used in engineering geological survey because of the advantages of fast and contactless measurement and acquisition,but these two methods are greatly influenced by the terrain and environment,and there are easily blind areas for acquisition,while UAV photogrammetry technology is efficient,easy to operate and less influenced by the environment,and can quickly establish 3D realistic models and3 D point cloud models of rocky slopes,which is convenient for structural surface identification and geometric information extraction.information extraction.In this thesis,the research object is to extract 3D point cloud data of rocky slope by UAV photogrammetry,to conduct research on intelligent identification of rocky slope structure surface,division of dominant group and geometric information extraction,and to qualitatively analyze the stability of road rocky slope by combining with geological survey data,the specific work contents and results are as follows:(1)A new method of semi-automatic identification and grouping of rock structural surfaces based on 3D point cloud data is proposed,and the nearest node search(KNN)is used to construct the point cloud neighborhood set;the point cloud normal vector and curvature values are estimated by principal component analysis(PCA);the structural surface segmentation and dominant group division are carried out by using the improved area growth method,with the point with the smallest curvature as the starting growth point,the point cloud normal vector angle and curvature as the The growth condition is set as the normal vector angle and curvature of the point cloud,and a threshold value is set to segment a single structural surface.After obtaining all the structural surface data,the normal vector angle of the structural surface is used as the growth condition and the threshold value is set to cluster the dominant group of structural surfaces;the random sampling consistency algorithm(RANSAC)is used to fit all the structural surfaces,extract the best plane and calculate the yield and trace length information based on the plane normal vector and point cloud coordinates.(2)The structural surface identification,geometric information extraction and dominant group division are carried out for typical road rocky slope test data,and the results are compared and analyzed with those of other methods,which show that the structural surface identification,geometric information extraction and dominant group division of this paper are basically consistent with other methods.The average difference between the inclination and dip angle of the structure surface is 3.24° and 2.74°,which is better than 3.87° and 3.03° of other methods,and the medium error of the inclination and dip angle is 4.54° and 3.58°,which meets the requirement of tolerance error of 12° and9°.(3)Using UAV photography technology,planning data acquisition plan and setting reasonable parameters,collecting high-definition 2D images of rocky slopes in Banan District,Chongqing,combining tilt photogrammetry and close photogrammetry to establish high-quality 3D real-world models and extract 3D point cloud data.(4)Based on the point cloud data of Jialong Road rock slope in Banan District of Chongqing,three identification regions were divided,and the occurrence and trace information of structural plane and its dominant groups were extracted by using the method presented in this thesis.The distribution map and density map of occurrence poles were drawn to show the spatial distribution of structural plane and judge the rationality of the division of dominant groups.The vertical projection map of structural plane of dominant group was drawn and the unstable region and failure form of rock slope were identified by combining slope inclination and slope Angle.The trace length distribution range and the development of structural plane were determined by drawing the trace length and trace frequency map.