Image quality assessment using natural image statistics in gradient domain

It is a challenging work to find an efficient metric of image quality assessment, which is an important approach for many image processing tasks. In this paper, we propose a novel reduced reference image quality assessment based on natural image statistical prior in gradient domain. The derivatives of image describe the detailed geometric features of images which are appealing for visual prediction of human beings. Resistor average distance as a symmetric dissimilarity measure between distributions of images is utilized in the definition of distortion measure. Our method is general purpose for different distortion types and has relatively low complexity. The experimental results upon image database show that the proposed method is consistent with the subjective assessment of human beings and has a good performance for all distortion types.