Fast Digital Shearlet Transform Based Image Quality Assessment

Abstract

In this paper, a new image quality assessment based on structural similarity in shearlet transform domain is presented. The fast digital shearlet transform is based on the pseudo-polar fourier transform and can represent the image at different scales and directions. The correlation indexes between the reference image and distortion image are calculated in every subbands and directions. The final quality of distortion image can be obtained from weighted sums of all subbands. Experimental results show that the proposed method has a good consistency with the subjective assessment of human beings and improves accuracy of image quality prediction.