A geometric compression algorithm for massive terrain data using delaunay triangulation
Soo Kyun Kim · 1999
In this paper we introduce a new compression technique for a large triangulated terrains using Delaunay triangulation. Our compression technique decomposes a triangulated mesh into two parts. One is a point set whose connecting structure is defined implicitly by Delaunay edges. The other is the set of edges which cannot be recovered by the implicit Delaunay triangulation rule. Thus we only need to keep the whole vertex coordinates and a few edges which is not included in the Delaunay edges. For the vertex coordinate, we apply "entropy coding" given by [Costa98], and we store only the edges not included in Delaunay triangulation. In experiments, we prepared several TIN data set with various resolutions, which were generated by five typical algorithms for terrain simplification. Those algorithms include progressive meshing, vertex decimation and incremental greedy insertion etc. We found that most of terrain triangulations are quite similar( = nearly 93%) to the plane-projected Delaunay ...