DESIGN OF GENERALIZED FRAMEWORK FOR NO REFERENCE COLOR IMAGE QUALITY ASSESSMENT

Abstract

proposed work presents a generalized newlineframework for no-reference color image quality assessment based on singular value statistics newlineusing Laplacian pyramid decomposition of the image. Besides being simpler and faster the newlinemetric shows correlation with human perception while predicting image quality for different newlinedistortions. It makes an attempt to reduce computational burden by using Laplacian pyramid newlinedecomposition instead of image transform based coding. In future, in order to increase the newlineaccuracy of prediction, image quality evaluation can be preceded by an algorithm for newlineidentification of distortion. The proposed metric can also be modified in order to extend it to newlinevideo quality assessment applications. newline