A data compression method on chinese fonts using classification techniques

摘要

A new coding method is presented to reduce the storage requirement of the contours of Chinese characters so that a high compression ratio of Chinese fonts can be reached. Using the classification techniques, the similarity among the contours of Chinese characters can be measured; if two contours have a very high similarity, one of the two contours will be deleted to reach a higher compression ratio. The present method is tested experimentally on a set of 226,724 Chinese contours of three style fonts and has the following advantages: (1) high compression ratio capability, (2) high encoding speed, (3) simple and quick decoding algorithm, (4) able to compress the contours with multi-resolution, (5) the quality of the generated Chinese fonts is almost the same as that of the uncompressed Chinese fonts, (6) able to use some error-free compression techniques to re-compress the storage requirement compressed by the present method. The presented technique can be executed on most of the low-cost microcomputers, and has already been applied to improve the practical products of the Fast Chinese Character Zooming and Generating System.