Efficient approach of image quality evaluation based on fuzzy logic

Abstract

The image quality evaluation is an important subject in image processing techniques. Aiming to improve the existing methods of image quality measurement, a new fuzzy logic-based image quality evaluation approach is presented in this paper. It is well known that an image is composed of details (edge, texture) and smooth regions, according to several fuzzy inference rules, three different quality factors are defined and applied to their respective regions. The proposed approach separates the distorted effects from the original edges in the image, and avoids the edges being mistaken as distorted effects. Experimental results show that the proposed method is robust for various image, in addition, it has the general performance of image quality evaluation approaches. © 2013 Springer-Verlag.