A new image quality measure

Measuring image quality is an interesting and challenging area of research. In this paper we investigate the performance of the statistical functions called copula as image quality measures. These functions are popular for applications where data distributions are unknown. This property motivated some researchers to using these copulas in image processing in general and in detecting image changes and image registration in particular. In this research, we use the Gaussian copula to calculate the mutual information, which is the measure of the association of the reference and the distorted or tampered with images. To test the performance of the proposed method, we implemented our method on LIVE image database and compared our results with three popular image quality measures namely Visual Information Fidelity (VIF), Structural Similarity (SSIM), and Universal Quality Measure (UQI). The results show that our quality measure, obtained similar results to the three methods in 99% of the time, hence the proposed method can be considered as an efficient image quality index.