A Novel Discrete Wavelet Domain Error-Based Image Quality Metric with Enhanced Perceptual Performance
Soroosh Rezazadeh, Stéphane Coulombe · International Journal of Computer and Electrical Engineering · 2012
In this paper, we present an error-based method to compute image quality scores in the discrete wavelet domain using the Haar wavelet.The proposed method does not extract any human visual system parameters, which noticeably reduces its complexity, and relies on computation of the errors (differences) at an appropriate level of wavelet decomposition.We also propose a formula to automatically calculate the required level of wavelet decomposition at a desired viewing distance.To consider the effect of very fine image details in quality assessment, the proposed method defines a multi-level edge map for each image which comprises only the most informative image subbands.We present the mathematical complexity analysis of our method and compare it to the conventional PSNR.We show that when the defined edge map is not taken into account we can calculate the quality score with lower complexity than with the PSNR, and with a prediction accuracy close to that of SSIM.We tested the performance of our method on two different databases: an image database and a video database.