Image Compression on Hierarchical Decomposition and Pixel Prediction

This paper presents lossless color image compression algorithm based on the proposed method, "Hierarchical decomposition and pixel prediction". An input RGB color image is transformed into YCuCv color space by an Reversible Color Transform (RCT). The luminance channel is encoded by any of the lossless image compression such as Context Based Adaptive Lossless Image Codec (CALIC), JPEG-LS, or JPEG 2000 lossless. The chrominance channel Cu and Cv are encoded using hierarchical decomposition & pixel prediction. Arithmetic coder is used to encode the prediction error. For several set of test images, this algorithm is shown that the proposed method is better than JPEG2000 and JPEG-XR.