3D Modeling Technology Of Open Space Of Landslide Body And Analysis Of Terrain Characteristics Change In Xiaojiang River Basi

Landslides and mudslides are geological hazards commonly found in the mountainous areas of south-western China,and are characterised by their suddenness and danger,which not only cause serious casualties but also pose a great threat to the development of the national economy.It is therefore a very important task in the field of geological hazard research to monitor and analyse changes in landslides and to provide timely warning of landslide disasters so that preventive measures can be taken.With the rapid development of measuring instruments and remote sensing technology,UAV and 3D laser scanning technology are widely used in the field of landslide change monitoring with the advantages of non-contact,high efficiency and high accuracy.This paper takes a landslide in the Xiaojiang river basin of Dongchuan district,Kunming as the research object,and combines UAV technology and 3D laser scanning technology to acquire two phases of data from the study area.Finally,the landslide is analysed in terms of terrain factors,characteristic terrain elements and point cloud models to achieve multi-directional analysis of landslide features.Through the experiments,the main findings of this paper are as follows:(1)To investigate the universality of three filtering algorithms,IPTDF,CSF and SBF,for Li DAR point cloud data processing of complex terrain in debris flow gullies;when the terrain is flat and the slope is small,the average total errors of IPTDF,CSF and SBF algorithms are 8.73%,8.81% and 8.90% respectively,which are very close in value,and all three algorithms can achieve the purpose of classifying ground points,while When the terrain is complex and the slope is large,the average total errors of IPTDF,CSF and SBF algorithms are 13.4%,36.4% and 30.2% respectively,CSF and SBF algorithms have lower accuracy in the steep-walled area with complex terrain,while the average total error of IPTDF algorithm is much lower than the other two algorithms,which can be well adapted to the complex geomorphological characteristics of the mudslide gully;study the IPTDF algorithm The applicability of the IPTDF algorithm to unmanned point clouds of landslides is investigated,and the results show that the method extracts ground points well,and for ridges and some minor terrain features are well preserved,which achieves efficient and accurate classification of ground points.(2)Combining the unified coordinate system method and ICP algorithm to complete the fusion of point cloud data,RMSE=0.113m;the fused point cloud data is used to construct a 3D model of the landslide body based on NURBS surface,and the final model constructed shows that the multi-source point cloud data is feasible for landslide body model construction,and the surface of the model is relatively smooth and retains terrain features,but there is also some loss of texture details.(3)Six interpolation methods,namely NN,Kriging,IDW,Spline,Trend and Topo to Raster,were used to construct the landslide DEMs,and the accuracy of the DEMs constructed by each method was evaluated using the cross-validation method.The DEM constructed by Kriging has the lowest root mean square error and the lowest mean absolute error.The DEM constructed by Kriging has a more accurate description of the local details of the mountain body and can be well applied to the interpolation of point cloud data in complex terrain;the DEM constructed by Kriging interpolation method extracts topographic factors such as slope,slope direction and topographic relief and characteristic topographic elements,such as characteristic points and characteristic lines.(4)The analysis of the changes in surface characteristics of the landslide body was carried out from three aspects: changes in topographic factors,changes in characteristic terrain elements and changes in the model.The landslide body tended to stabilize during the period from 2022.06.01 to 2022.11.05.There were no obvious changes in the elevation of the point cloud in the two phases of the landslide body,and landslides existed in some areas,and mounding uplift was formed in several valleys;changes in slope direction were relatively small,with changes in slope The change in slope direction is concentrated in the north,northeast and southwest directions;the change in the point cloud of the landslide body in the X and Z directions is greater than the change in the Y direction,and there is no large-scale landslide disaster on the landslide body as a whole.