A Novel Data Compression Scheme for Chinese Character Patterns

Abstract

This paper presents and efficient lossless compression scheme for Chinese character patterns. The scheme analyzes the characteristics of line-Segment structures of Chinese character patterns. A novel matching algorithm is developed for the line-segment prediction which is used in the encoding and decoding processes. The bit rates achieved with the proposed lossless scheme are 02653, 0.2583, for three Chinese fonts, respectively. Due to the black and white points in Chinese character patterns are highly correlated, subsampling and interpolation schemes are considered to further increase the compression ratio. With these schemes, the low bit rates are achieved. Compared to the available algorithm, our approach gives the better performance. Two types of interpolation techniques are presented for the enlargement of Chinese character patterns. They are the 2-D interpolation and spline interpolation. Compared with the lossless compression scheme, the 2-D subsampling scheme can further reduce the bit rates as high as 43.19%, 41.83% for three widely used Chinese fonts. Respectively.