Recently,Geomorphological remote sensing research team of School of Surveying and Mapping Science and Technology, Xi'an University of Science and TechnologyThe research team of mine intelligent surveying and mapping, together with the Yellow River Soil and Water Conservation Suide Governance and Supervision Bureau (Suide Soil and Water Conservation Scientific Experimental Station), the Yellow River Soil and Water Conservation Ecological Environment Monitoring Center, Iran's KN Tush University of Technology and other units, published the research results entitled "Remote Sensing of Environment", an international top journal in the field of remote sensing.
The first author of this ** is Yang Xin, a master's student of the School of Surveying and Mapping, and the corresponding authors are Professor Li Pengfei and Professor Tang Fuquan. RSE journal (TOP in Zone 1, Chinese Academy of Sciences, IF=13.)5) It is recognized as the most authoritative journal in the field of remote sensing, ranking first among similar SCI academic journals in the world for a long time, mainly publishing original articles related to the theory, application and method of earth remote sensing observation.
The application of 3D laser point cloud data in the fields of soil erosion, land surface subsidence and river landform evolution is becoming more and more extensive, and how to accurately quantify the changes of complex 3D terrain and landform is still a difficult problem in current geoscience research.
At present, there are 1D terrain deformation distance quantization algorithms (C2C, C2M and M3C2 algorithms) that cannot directly quantify the terrain change volume,2The 5D Terrain Change Volume Quantization Algorithm (DOD Algorithm) cannot effectively distinguish between positive and negative terrain changes in complex 3D terrain scenes, and the problem of terrain surface overlap will occur in the process of constructing DEM, while the 3D algorithm (3D-M3C2 algorithm) solves the above problems, but the algorithm is still severely limited by the point cloud density and point cloud morphology, and the monitoring accuracy is low in the terrain scene where the point clouds are sparse and the point cloud morphology is different. In view of this, this study proposes a volumetric slicing algorithm (SCCD) based on the slicing idea and the Laplace shrinkage principle.
The algorithm can accurately obtain the contour of point cloud slices and effectively eliminate the uncertainty of topographic point clouds, which overcomes the limitations of traditional volumetric slicing algorithms that are difficult to be applied to complex 3D terrain change monitoring. The SCCD algorithm is compared with the existing terrain change volume quantization algorithms (3D-M3C2 and DOD algorithms) in terms of terrain deformation volume quantification accuracy, spatial distribution of terrain deformation, sensitivity to topographic point cloud density change and sensitivity to morphological difference of topographic point cloud.
The results show that the comprehensive monitoring ability of the SCCD algorithm is better than that of the 3D-M3C2 and DOD algorithms, which is of great significance for the study of extremely complex terrain changes. In addition, the contour of the point cloud slice obtained by the SCCD algorithm also provides a quantitative basis for the study of complex 3D terrain parameters.