Computing Farthest Neighbors on a Convex Polytope
Let N be a set of n points in convex position in ∝3. The farthest-point Voronoi diagram of N partitions ∝3 into n convex cells. We consider the intersection G(N) of the diagram with the boundary of the convex hull of N. We give an algorithm that computes an implicit representation of G(N) in expected O(n log2 n) time. More precisely, we compute the combinatorial structure of G(N), the coordinates of its vertices, and the equation of the plane de?ning each edge of G(N). The algorithm allows us to solve the all-pairs farthest neighbor problem for N in expected time O(n log2 n), and to perform farthest-neighbor queries on N in O(log2 n) time with high probability. This can be applied to find a Euclidean maximum spanning tree and a diameter 2-clustering of N in expected O(n log4 n) time.
KeywordsVoronoi Diagram Voronoi Cell Implicit Representation Convex Polytope Connected Subgraph
Unable to display preview. Download preview PDF.
- 3.Sergei N. Bespamyatnikh An efficient algorithm for the three-dimensional diameter problem. In Proc. 9th ACM-SIAM Sympos. Discrete Algorithms, pages 137–146, 1998.Google Scholar
- 5.K. Mulmuley Computational Geometry: An Introduction Through Randomized Algorithms. Prentice Hall, Englewood Cliffs, NJ, 1994.Google Scholar
- 6.F.P. Preparata and M.I. Shamos Computational Geometry: An Introduction. Springer-Verlag, New York, NY, 1985.Google Scholar
- 7.E. Ramos Construction of 1-d lower envelopes and applications. In Proc. 13th Annu. ACM Sympos. Comput. Geom., pp. 57–66, 1997.Google Scholar
- 8.E. Ramos An optimal deterministic algorithm for computing the diameter of a 3-d point set. In Proc. 16th Annu. ACM Sympos. Comput. Geom., page to appear, 2000.Google Scholar