Colour image compression based on the measure of just noticeable colour difference

To the human vision, there exists in colour images a certain amount of perceptual redundancy since the human visual system (HVS) has limited sensitivity in discriminating colour signals of small differences. By measuring the perceptual redundancy inherent in colour images and shaping the coding distortion into the perceptual redundancy, colour images are expected to be represented more efficiently. Approaches to perceptually optimise the efficiency of image coders in compressing colour images with the perceptual redundancy estimated by a colour visual model are presented. The model estimates the perceptual redundancy for each colour pixel as a visibility threshold of colour difference in any colour space and in a spatial or frequency domain. Two existing image coders are modified to take advantage of the perceptual redundancy and simulated to inspect if their coding efficiency is improved. In the spatial domain, the JPEG-LS coder in the near-lossless compression mode is modified to make coding errors part of the perceptual redundancy in compressing colour images in the RGB space. In the wavelet domain, the JPEG2000 coder is refined by minimising the perceptible distortion involved in the rate control of the compressed image in the YCbCr space. Simulation results show that, in both cases, the performance of the perceptually tuned coder is superior to that of the un-tuned coder in terms of the bit rate required for achieving the same visual quality.