Refine your search
Collections
Year
A B C D E F G H I J K L M N O P Q R S T U V W X Y Z All
This paper aims to combine the effects of blurring and blocking on a JPEG image to devise a quality metric that correlates well will human vision. Blurring is caused due to the degradation of the high frequency components in an image. Blocking is another irregularity that is present in images that use block based coding. It is also called as quilting or checker boarding. Both these artifacts greatly reduce quality of the images. Blurring is estimated using multi-scale quadrature filters and blocking is measured using perceptual blockiness method. Both the blurring and blockiness estimation methods are combined using a weighted approach based on thresholding.