离散地形曲面上的精确Voronoi图直接生成算法

Direct Algorithm for the Exact Voronoi Diagram on Discrete Topographic Space

  • 摘要: Voronoi图是地学计算中的一个基本结构,但是在地形曲面上,它还缺乏能与平面Voronoi图媲美的精度和成熟的算法。在离散地形曲面的不规则三角网格网上引入计算几何的测地距离场,从格网边上的距离场奇点逐步生长代表平分线的双曲线,由双曲线的排列得到离散曲面的精确划分,再将划分的面片聚类, 生成精确的测地Voronoi图(geodesic Voronoi diagram, GVD)。然后,从定量与定性两方面对精确Voronoi图进行了检验,证明GVD可以给地形曲面空间分析带来基础性改进。基于奇点生长和双曲线排列的直接算法避免了现有算法对格网面片的过度细分与预处理,整体上直观易行,为数字地形分析发展严密的Voronoi图分析提供了有益探索。

     

    Abstract:
      Objectives  Voronoi diagram is a fundamental structure in geo-computing, but it still encounters the problem of exactness and the challenge of an exact algorithm comparable to planar Voronoi diagrams in the topographic space.
      Methods  The geodesic distance field of computational geometry is introduced into the triangulated irregular network in the discrete topographic surface. The hyperbolic curves representing the bisector are gradually grown from the singularity of the distance field on the edge of the grid. The precise division of the discrete surface is obtained by the arrangement of the hyperbolic curves, and the exact geodesic Voronoi diagram (GVD) is obtained by clustering the divided patches. Then, the exact Voronoi diagrams are tested quantitatively and qualitatively.
      Results and Conclusions  It is found that the exact GVD can bring a basic improvement for the spatial analysis of topographic surface. The direct algorithm based on singular growth and hyperbolic arrangement is intuitive and easy to implement, which avoids the excessive subdivision and preprocessing of grid patches by the existing algorithms, and provides a useful exploration for the development of Voronoi diagram analysis in digital topographic analysis.

     

/

返回文章
返回