Study on image quality assessment methods based on human visual sensitivity

Image quality assessment study becomes one of the crucial basic topics in image information. We propose a new approach for image quality assessment, which indicates the image quality synthetically with kinds of factors such as image ambiguity, correlation and luminance extracted by wavelet transforms. Compared with the traditional criteria for image quality, the experiment results show our method can achieve a more comprehensive image quality assessment performance, and fit for human visual sensitivity much better.