Prediction error based compression of color images using WDR coding

This work presents an efficient compression of color images using Wavelet Difference Reduction (WDR) coding and prediction error quantization. After reducing the correlation among RGB planes using RGB to YCbCr transform, each of the transformed planes are decomposed into several sub-images. The sub-images of Y plane are compressed independently without much more degrading the image quality, while each of the sub-images of Cb and Cr planes are compressed using quantization and WDR and Huffman coding techniques. At the receiver side, the original image content is recovered with the help of compressed data and quantizing parameters, used during compression. To give the evidence of our findings the proposed work is compared with several wavelet based compression techniques. The experimental results and analysis of the proposed work demonstrate its superiority both in terms of objective and subjective fidelity criteria.