A Haar wavelet approach to compressed image quality measurement

The traditional mean-squared-error and peak-si,anal-to-noise-ratio error measures are mainly focused on the pixel-by-pixel difference between the original and compressed images. Such metrics are improper for subjective quality assessment, since human perception is very sensitive to specific correlations between adjacent pixels. In this work, we explore the Haar wavelet to model the space-frequency localization property of human visual system (HVS) responses. It is shown that the physical contrast in different resolutions can be easily represented in terms of wavelet coefficients. By analyzing and modeling several visual mechanisms of the HVS with the Haar transform, we develop a new subjective fidelity measure which is more consistent with human observation experience. (C) 2000 Academic Press.