Image quality assessment based on noise classification

Although the structural information-based image quality assessment SSIM (Structural Similarity) is simple and has been proved to be better than the PSNR (peak signal to noise ratio) method, there are still some difficulties in assessing various noise images. Considering the effect of the Gaussian and Salt & Pepper noise on image quality, this paper propose a two-step strategy for image quality assessment based on Noise Classification (NC-SSIM). The novel method firstly classifies noise types based on flat regions, then improve existing SSIM algorithm. We test the validation of the proposed algorithm on particular subsets of new TID2008 database, and the experiments show that the NC-SSIM model can assess the image quality more precisely than the SSIM method.