Unified lossless and near-lossless color image coding based on adaptive quantization

This paper proposes a unified coding algorithm for lossless and near-lossless compression of still color images. The algorithm can control the Peak Signal-to-Noise Ratio (PSNR) of the reconstructed image. Furthermore, the distortion on the RGB plane is suppressed to within the level of /spl plusmn/p, where p is a certain small non-negative integer. In order to control the PSNR of the reconstructed image, an adaptive quantizer is designed on the basis of human visual criteria. Experimental results confirm the effectiveness of the proposed algorithm.