Implementation of a randomized algorithm for Delaunay and regular triangulations in three dimensions

https://doi.org/10.1016/0167-8396(94)00018-NGet rights and content

Abstract

For a set of n points in R3, we can define the Delaunay triangulation of the point set. For each assignment of weights to the points there is a regular triangulation of the point set. We describe the implementation of algorithms to compute the Delaunay triangulation of an unweighted point set and the regular triangulation of a weighted point set. The algorithms run in expected time O(n log n) for uniformly distributed points and other dense point sets. Worst case point sets, which do not occur very often in practice, require O(n2) expected time. The software described in this paper is available via anonymous ftp at ftp.ncsa.uiuc.edu.

References (17)

  • B. Joe

    Construction of three dimensional triangulations using local transformations

    Computer Aided Geometric Design

    (1991)
  • C.L. Lawson

    Software for C1 surface interpolation

  • F. Aurenhammer

    Power diagrams: properties, algorithms, and applications

    SIAM J. Comput.

    (1987)
  • T.J. Barth et al.

    Three-dimensional unstructured grid generation via an incremental insertion and local optimization

  • J.C. Cavendish et al.

    An approach to automatic three-dimensional finite element mesh generation

    Internat. J. Numer. Meth. Eng.

    (1985)
  • B.N. Delaunay

    Sur la sphère vide, Izv. Akad.

    Nauk SSSR, Otdelenie Matematicheskii i Estestvennyka Nauk

    (1934)
  • D.P. Dobkin et al.

    Primitives for the manipulation of three-dimensional subdivisions

    Algorithmica

    (1989)
  • H. Edelsbrunner

    Algorithms in Combinatorial Geometry

    (1987)
There are more references available in the full text version of this article.

Cited by (41)

  • A triangulation-based approach to automatically repair GIS polygons

    2014, Computers and Geosciences
    Citation Excerpt :

    The fact that we start from the outside is key to ensuring that the algorithm performs correctly. To find a triangle located outside any ring, we exploit the “far-away point” (also called the “big triangle”) that is used by several CT implementations (Liu and Snoeyink, 2005; Facello, 1995). In brief, every edge on the border of the convex hull has a triangle incident to it, and this triangle is formed by the edge and a special “infinite” point.

  • Lipid-binding surfaces of membrane proteins: Evidence from evolutionary and structural analysis

    2011, Biochimica et Biophysica Acta - Biomembranes
    Citation Excerpt :

    VOLBL uses weighted Delaunay triangulation and alpha shape to compute metric properties of molecules. The Delaunay triangulation of membrane proteins was computed using the DELCX program [24,25], and the alpha shape was computed using the MKALF program [24,26]. The van der Waals radii of protein atoms were from Tsai et al. [27].

  • Adaptive tetrahedral element generation and refinement to improve the quality of bulk metal forming simulation

    2007, Finite Elements in Analysis and Design
    Citation Excerpt :

    The major advantage of the tetrahedral elements is the robustness in generating highly adaptive elements in an automatic manner for a complicated shape. The main stream of research on triangular/tetrahedral mesh generation has been centered around the Delaunay triangulation method [37–45]. In recent years, remarkable advances have been made in mesh quality control and/or optimization [46–51].

  • Sliver-free perturbation for the Delaunay tetrahedrization

    2007, CAD Computer Aided Design
    Citation Excerpt :

    In this sense, the Delaunay tetrahedrization for integer-grid points cannot be perturbed without generating slivers. On the other hand, it has been pointed out in several works [4,11,14,17] that a sliver-free perturbation is possible if we reduce the problem into the corresponding convex hull problem in one higher dimensional space by the lift-up technique and perturb the points in the higher-dimensional space. This perturbation can be interpreted in the original space as the perturbation of the radii of the spheres in the Voronoi diagrams for spheres; originally, the radii of the spheres are zero and they are perturbed into infinitesimally small positive radii [4,17,22].

View all citing articles on Scopus

Present address: 2432 Digital Computer Lab., 1304 West Springfield Avenue, Urbana, IL 61801, USA.

View full text