Shape data compression method, shape data decompression method, shape

Abstract

A shape data compression method for image generation by three-dimensional computer graphics, including a first stage inputting a code number and a quantization precision for encoding a three-dimensional coordinate point sequence and a two-dimensional coordinate point sequence, and reading the coordinate point sequence data. A second stage includes mapping transformed point sequences in a one-dimensional space. A third stage calculates an initial apace division width based on quantization precision, divides the normal space into partial spaces and analyzes distribution of the transformed point sequences therein, decides a division width and calculates distribution of transformed point sequences in each partial space at the division width. A fourth stage obtains a mean value of coordinate values of transformed point sequences distributed in the partial space, and encodes them to produce a code book. A fifth stage generates code sequences for encoding the transformed point sequences according to the code book. And a sixth stage outputs compressed data comprising the transformation parameters for data mapping in the second stage, the code book, and the code sequences.