A LOSSLESS COLOR IMAGE COMPRESSION USING AN IMPROVED REVERSIBLE COLOR TRANSFORM WITH HIERARCHICAL PREDICTION AND MODIFIED HUFFMAN CODING

Abstract

In case of the conventional lossless color image compression methods, the pixels are interleaved from each color component, and they are predicted and finally encoded. In this paper, we propose a lossless color image compression method using hierarchical prediction of chrominance channel pixels and encoded with modified Huffman coding. An input image is chosen and the R, G and B color channel is transform into YCuCv color space using an improved reversible color transform. After that a conventional lossless image coder like CALIC is used to compress the luminance channel Y. The chrominance channel Cu and Cv are encoded with hierarchical decomposition and directional prediction. The effective context modeling for prediction residual is adopted finally. It is seen from the experimental result the proposed method improves the compression performance than the existing method.