Inter-Channel Correlation Based Color Image Compression

Most of the color image compression techniques reduce the redundancy between color components (R, G, B) by transforming the color primaries into a de-correlated color space, such as YIQ or YUV. As the human visual system is more sensitive to details in luminance than to details in chrominance, the chrominance components can be compressed at high rate. Instead of de-correlating the color planes, high regional correlation between the components of color images is used as a basis for a new coding technique. The high correlation of color channels implicitly suggests a localized functional relation between the components. It could be used in an alternative compression approach, by approximating subordinate colors as functions of the base color instead of coding each color planes. The 9 by 7 wavelet filter and Set Partitioning in Hierarchical Tree (SPIHT) coding for encoding the base color are used. The linear three channel Discrete Cosine Transform (DCT) is applied to the RGB image to obtain the color channels C1C2C3. Taking C1 as base color and the other two channels are approximated as linear functions of the base color, only two parameters are transmitted for each block N*N of C2 and C3. A significant Peak Signal to Noise Ratio improvement is achieved compared to the traditional coding scheme for the same compression rate. The proposed algorithm allows the reduction of complexity for both coding and decoding of color images.