Color Image Compression Coding Using the Correlation of Color Components
利用色彩分量相关性彩色图象压缩编码

This paper presents a new method for color image compression coding by using correlation of color components combined with wavelet transform. In this method, one of the three color components of a color image is coded using wavelet transform zerotree coding, and the component is segmented into several non overlapping subregions in the same size. and the chromatism matrixs are figured out by the margin between the component and the other two components in each subregions, and then the correlation coefficients is ciphered out suing chromatism matrix analysis. Based on the calculated correlation coefficients, the coding for the other two components becomes the coding of correlation coefficients. During decoding, the component is first reconstructed using inverse quantization and wavelet inverse transform. Then according to this component and the reconstructed correlation coefficients, the other two components are calculated out. Finally the color image is reconstructed by the three color components. Experiment results show that this method is of high coding and decoding rates, and with quite good signal to noise ratio, compression rate and visual quality.