Low bit rate image coding based on wavelet transform and color correlative coding

Along with the rapid development of information processing, people attach more and more importance to image compression. Most coding techniques for color image compression employ a de-correlation approach, the RGB space is transformed into a de-correlation color space, then the de-correlated color components are encode separately by the same method for examples JPEG and JPEG2000. A different image encoding method of correlation and wavelet transform approach (CWA) based on human visual system, is presented in this paper. Taking into account the human visual characteristics and instead of de-correlating color components, we employ the existing-color correlation Y, G, R components, after Wavelet transform of these components, G component Wavelet Coefficient is the base component to approximate Y, R components' Wavelet Coefficient on the basis of the least linear squares. Then in order to enhance the algorithm's performance, when the Wavelet Coefficient approximation of Y, R components, different sub-band choice different size block on different sub-band characteristics. G component Wavelet Coefficients after of quantization is encoded using adaptive Huffman coder based on different sub-band, and Huffman bit stream can be encoded using arithmetic coder. The Wavelet Coefficient approximation of Y, R components convert char stream, then is encoded using Huffman coder. Experimental results show that the proposed correlation and wavelet transform image coder based on human visual system offers coding performance superior to presently available algorithms based on the common de-correlation approach.