Image quality measurement using the Haar wavelet

The traditional mean-squared-error (MSE) or peak-signal-to-noise-ratio (PSNR) error measures are mainly focused on the pixel-by-pixel diierence between the original and compressed images. Such metrics are improper for subjective quality or delity assessment, since human perception is very sensitive to correlations between adjacent pixels. In this work, we explore the Haar wavelet to model the space-frequency localization property of human visual system (HVS). It is shown that the physical contrast in diierent resolutions can be easily represented in terms of transform coeecients. We model HVS with the Haar lter with several visual mechanisms and develop a subjective quality measure which is more consistent with human observation experience.