A visual model for subband image coding

Abstract

In this paper, we propose a visual model for the general subband coding system. This model is a generalization of Watson's visual model for DCT. Based on this model, a new perceptual distortion is proposed which can be used together with an efficient bit allocation algorithm to determine the quantizer stepsizes of subband coders. Simulation results and subjective viewing tests demonstrate that the proposed approach is very effective in improving the subjective quality of the encoded image at low bit rate.