顾及地形断裂线的高保真DEM重建研究  

search on High Fidelity DEM Reconstruction Considering Terrain Fracture Lines

作　　者：夏袁菡子 周胜洁 邸国辉 XIA Yuanhanzi;ZHOU Shengjie;DI Guohui(Hubei Institute ofWater Resources Survey and Design Co.,Ltd.,Wuhan 430070,China)

机构地区：[1]湖北省水利水电规划勘测设计院有限公司,湖北武汉430070

出　　处：《地理空间信息》2025年第2期20-24,共5页Geospatial Information

基　　金：湖北省水利重点科研项目(HBSLKY202112)。

摘　　要：数字高程模型(DEM)是目前最重要的地球信息三维基础产品之一,但在植被茂盛等造成地形突变处点云密度低的情形下,生成的DEM局部误差偏大,出现不同程度的失真。针对DEM生产过程中的突变地形,基于法向量法来提取断裂线,提出了以虚拟断裂线法或拓扑断裂线法来表达断裂线。实验分别利用克里金插值法和径向基函数法来重建DEM,通过比较LiDAR脚点缺失、虚拟断裂线少点、虚拟断裂线密集点以及拓扑断裂线法4个不同组别所产生的高程误差,得到如下结果:引入拓扑断裂线法后,克里金法生成的DEM中误差为0.14 m,比点云缺失时中误差减少0.511 m,径向基函数法生成的DEM中误差为0.13 m,比点云缺失时中误差减少0.515 m,实验效果表明在点云密度不足的情况下,拓扑断裂线法在2种插值算法中均适用。Digital elevation model(DEM)is currently one of the most important 3D basic products for Earth information.However,in situations where point cloud density is low caused by lush vegetation and other factors,the generated DEM has a large local error,resulting in varying degrees of distortion.We focused on the sudden terrain changes in the DEM production process,used the normal vector method to extract fault lines,and proposed using virtual fault line method or topological fault line method to express fault lines.We used Kriging interpolation method and radial basis function method to reconstruct DEM,and compared the elevation errors generated by four different groups of LiDAR,such as missing foot points,few virtual fault lines,dense virtual fault lines,and topological fault line method.The results show that after introducing the topological fault line method,the mean square error of DEM generated by Kriging method is 0.14 m,which is 0.511 m less than that generated by point cloud missing.The mean square error of DEM generated by the radial basis function method is 0.13 m,which is 0.515 m less than that generated by point cloud missing.The experimental result indicates that the topological fracture line method is suitable for both interpolation algorithms in the case of insufficient point cloud density.

关 键 词：LIDAR 断裂线 DEM 克里金插值法 径向基函数法 

分 类 号：P284[天文地球—地图制图学与地理信息工程]