Image Quality Measurement by SVD Method and Two-Sample Kolmogorov-Smirnov Test

Abstract

Image quality assessment plays a significant role in image processing applications, such as control and enhances the quality of images. Objective image quality assessment attempted to evaluate the visibility of differences between distorted and original images and it is worthwhile, if it is close to human visual system. To this aim, we present a structural information-based image quality assessment algorithm, based on SVD factorization and Two-sample Kolmogorov-Smirnov test. This algorithm uses the SVD decomposition of each reference and distorted images and to make a comparison, it uses Two-sample Kolmogorov-Smirnov test. Some tests were applied to appraise the performance.