Image quality assessment is important in the image processing field. The traditional methods (e.g. PSNR and MSE) only compute the error between the reference image and distorted image at the pixel-scale

so they cannot well reflect the visual quality of image. Aiming at the above problem

a novel image quality evaluation approach is proposed according to the human visual characteristics. It is obtained to mimic human visual system and utilize similarity measure methods. Experimental results illustrate that the proposed has a good consistency with the subjective assessment of human beings; thus it can be used to describe the visual perception of the image effectively.