Image quality assessment based on complementary local feature extraction and quantification

Abstract

A novel full reference image quality assessment method based on statistical local representation from two complementary sources: log-Gabor wavelet representation and local derivative pattern is presented. The dissimilarity of these extracted features between distorted and reference images is quantified and mapped into an objective quality score. Experimental results on large-scale database show that the proposed method not only outperforms the state-of-the-art methods in terms of high accuracy of image quality prediction, but also is robust across different distortion types with high computation efficient.