Connectivity Coding: New Perspectives for Mesh Compression (Connectivity Coding: Neue Perspektiven für die Kompression von Dreiecksnetzen)
暂无分享,去创建一个
Compact encodings of the connectivity of planar triangulations is a very important subject not only in graph theory but also in computer graphics. For triangle meshes used in computer graphics the topologically planar regions dominate by far. New results by Isenburg et al. [6] even show that the connectivity is sufficient to describe shape by itself. Most coding methods for planar triangulations can be extended to manifolds of bounded genus with the same upper and lower bounds on the bit rate. In 1962 Tutte enumerated the number of different planar triangulations and his results show, that at least 3.245 bits per vertex are necessary to encode the connectivity graph of planar triangulations. In this article we improve the so far best upper bound [5] and show that the connectivity of a planar triangulation can be encoded with less than 3.525 bits per vertex, while ensuring a linear run time for encoding and decoding.
[1] Ian H. Witten,et al. Arithmetic coding for data compression , 1987, CACM.
[2] Jarek Rossignac,et al. Edgebreaker: Connectivity Compression for Triangle Meshes , 1999, IEEE Trans. Vis. Comput. Graph..
[3] Alon Itai,et al. Representation of graphs , 1982, Acta Informatica.
[4] W. T. Tutte,et al. A Census of Planar Triangulations , 1962, Canadian Journal of Mathematics.