A wavelet packet transform algorithm for color still images is presented and we show how and with what performances the transformed image tree can be pruned and the imagets altered to obtain a compression/decompression algorithm respectful of human psychovisual image perception. The basic assumptions for human vision on which the algorithm has been constructed are presented; and we deal with the color transformation used before applying wavelet packet transform. We also point out that a quasi-lossless compression/ decompression scheme can be easily obtained with a compression ratio up to 1:10 (quantization step was not considered here). Finally, a new quality criterion based on the results of tests on human sensitivity to various scale and color image details is proposed. This figure of merit can be considered a multiresolution release of the most commonly used peak SNR (PSNR) criterion.