Intelligent reorganized discrete cosine transform for reduced reference image quality assessment

Reduced reference image quality assessment does not require the presence of the original image for assessing the quality of a degraded image. This work proposes an intelligent method for reduced reference image quality assessment based on a reorganized discrete cosine transform (RDCT). A genetic algorithm (GA) is used to compute optimized estimation of the generalized Gaussian distribution (GGD), which then approximates the coefficient distribution in the RDCT domain. Experimental results validate that such an intelligent estimation produces far superior results compared to conventional empirical estimation methods as presented in the literature. We compare the proposed technique with a number of contemporary techniques in the literature and demonstrate the generalization capability and effectiveness of the proposed technique as compared to prior works.