Color image compression using adaptive color quantization

The objective of perceptual compression scheme is to remove the visual redundancy of images and maintain the perceptual quality of decoding result. In this paper, a perceptually adaptive color quantizer is designed for color image compression in wavelet domain. The quantization step sizes of three color components for each coefficient are adaptively adjusted by the estimated error visibility thresholds for achieving transparent coding at low bit rate. Based on the proposed color visual model, the error visibility thresholds are estimated by quantifying the perceptual redundancy in color spaces. The visual model is built on defining a perceptually indistinguishable region for each color in a color space by mapping colors, which are barely distinguishable from the target color in a perceptually uniform color space (PUCS), to the target color space. The simulation results show that the proposed scheme is able to achieve the perceptually lossless color image compression at low entropies.