Combined image similarity index

In the paper the idea of the combined image quality metric based on the structural and feature similarity comparison is discussed. Since most of image quality assessment methods developed during last years require the nonlinear mapping to obtain high correlation with subjective quality scores, there is an important problem of choosing the proper mapping function and its optimal parameters in practical applications. The most common approach is the utilization of the logistic function but its parameters strongly depend on the specific image quality database. To avoid the necessity of such mapping the nonlinear combination of three known image quality metrics is proposed and verified for seven currently available image quality assessment databases in terms of the linear correlation with subjective scores.