A distortion-based ranking of image quality indices using mutual information

Measurement of visual signal quality is of fundamental importance in a broad range of applications. The ultimate goal of quality assessment algorithms is to assess automatically the quality of images or videos in agreement with subjective human quality judgments. We discuss in this paper a new approach for ranking quality measures across different types of degradation that affect a given image. To rank the different image quality indices, we propose to use the concept of mutual information or information content. The experimental results show that the proposed ranking of quality indices is superior to the ranking based on second order correlation coefficients.