Extraction Of Typical Debris Flow Valley Terrain Features And Process Tracking Based On TLS Point Clou

The occurrence of debris flow disasters poses a threat to people’s lives and property safety.Debris flow disasters occur frequently in Dongchuan District,Yunnan.Debris flow valleys are one of the typical features of the region,which not only have the terrain characteristics of the valleys but also contain geological disaster information of debris flow landslides.It is of great practical and scientific significance to track and study the terrain of debris flow valleys and grasp their surface information and development patterns.3D laser scanning is an advanced technology for fully automatic and high-precision three-dimensional scanning of point clouds.The non-contact,high sampling rate,and high-precision characteristics of ground 3D laser scanners make them widely used in geological research.This article uses six years and five periods of data measured by a 3D laser scanner to construct a high-precision DEM using the obtained point cloud data,extract terrain indicators and important feature information,and track the development and changes of valleys.The main conclusions are as follows:(1)Compare five filtering methods,among which Bilateral filter has the highest ability to filter vegetation,moving surface filtering has the highest accuracy to separate ground points and non ground points,CSF ground filtering has the worst accuracy,moving surface filtering has a strong ability to filter high vegetation,and CSF ground filtering has a good filtering effect in areas with gentle slope,so combining the two methods,Using CSF ground filtering in gentle slope areas and moving surface filtering in other areas can achieve good filtering results.Compare the thinning effects of three point cloud thinning methods:spatial distance thinning,voxel segmentation thinning,and system(random)thinning at high(92%),medium(63%),and low(23%)sampling rates.The 63%spatial distance thinning method is the best choice.(2)Comparing the effects of Poisson surface reconstruction,NURBS surface reconstruction,and NNI,IDW,and Kriging interpolation methods in constructing DEM,according to the accuracy evaluation index calculation results,using the IDW method to construct DEM has the highest accuracy,with an average error ME of0.02505 m,a root mean square error RMSE of 0.08074 m,a fitting superiority R~2of0.99999,a fitting superiority closest to 1,and the best fitting effect.(3)Extract terrain indicators of debris flow valleys and obtain information on slope,aspect,ground undulation,and surface roughness.Introducing a method that combines hydrological analysis and slope aspect reclassification to extract high-precision ridge and valley lines.Considering the spatial coupling and differences of terrain feature points,four types of terrain feature points are extracted:ridge source points,ridge intersection points,runoff source points,and runoff nodes.(4)Qualitative and quantitative analysis is conducted on the changes in each period of the valley.The M3C2 distance algorithm is used to calculate the distance changes of point clouds.The differential DEM method is used to analyze the overall changes in the study area.Multiple period point cloud data are compared and analyzed for point,line,and surface features,and the process changes of surface details are obtained.Calculate the indicators of gully changes:the annual average projected area change rate of GA is 1199.81m~2/year,and GB is 446.753m~2/year;The average annual shrinkage rate of GA is 0.83m/year,and GB is 0.97m/year.Calculate the displacement changes of important terrain feature points in the X,Y,and Z directions over a period of six years.Among them,the point with the highest plane displacement within six years is the valley point sg4 and sg5,with a displacement distance exceeding 8m.The point with the highest elevation change is the valley point sg5,with a maximum increase of 10.909m.