计算机工程与应用 ›› 2011, Vol. 47 ›› Issue (36): 193-196.

• 图形、图像、模式识别 • 上一篇    下一篇

侵蚀坡面数字高程模型重构算法研究

龙满生1,徐洪志1,张道坤2,陈同聚1   

  1. 1.西北农林科技大学 信息工程学院,陕西 杨凌 712100
    2.西北农林科技大学 理学院,陕西 杨凌 712100
  • 收稿日期:1900-01-01 修回日期:1900-01-01 出版日期:2011-12-21 发布日期:2011-12-21

Building digital elevation model with eroded microtopography from point cloud

LONG Mansheng1,XU Hongzhi1,ZHANG Daokun2,CHEN Tongju1   

  1. 1.College of Information Engineering,Northwest A&F University,Yangling,Shaanxi 712100,China
    2.College of Science,Northwest A&F University,Yangling,Shaanxi 712100,China
  • Received:1900-01-01 Revised:1900-01-01 Online:2011-12-21 Published:2011-12-21

摘要: 构建具有微地貌的数字高程模型是坡面水蚀模拟的基础工作。针对定距邻域内插法的不足,提出了由点云构建侵蚀坡面的Delaunay三角内插法。该算法包括投影、三角化、分割和插值等步骤,利用主元分析法计算坡面点云的主轴和次轴并将其与坐标轴对齐,再由Delaunay三角化生成不规则三角网,将其分割为指定分辨率的规则格网,由三角线性插值得到格网单元的高程。试验结果表明,当邻域搜索半径与格网单元边长之比约为1.414时,定距邻域内插法的性能达到最优;当点云含有较多侵蚀坑等微地貌信息时,Delaunay三角内插法优于定距邻域内插法。

关键词: 点云处理, 数字高程模型, 主元分析, Delaunay三角化

Abstract: Building grid Digital Elevation Model(DEM) with microtopography is important for hydraulic erosion simulation on hillslope.Performance of the constant distance neighbor interpolation is analyzed and a Delaunay triangulation with linear interpolation is proposed to build the eroded slope surface from point cloud.This algorithm consists of projection,triangulation,segmentation and interpolation stages.Principal component analysis is used to compute the major and minor axes of the point cloud and to align them with the coordinate axes.A Triangulated Irregular Network(TIN) mesh is created by the Delaunay triangulation within the bounding box of the point cloud.The grid DEM is reconstructed by the linear interpolation inside the Delaunay triangle after the spatial domain is segmented into uniform grids with specified resolution.Experimental results show that the performance is optimal for the constant distance neighbor interpolation when the ratio of the searching radius to the grid resolution is about 1.414,and that the Delaunay triangulation interpolation is better than the constant distance neighbor interpolation in reconstructing the grid DEM with microtopography.

Key words: point cloud processing, digital elevation model, principal component analysis, Delaunay triangulation