No-Reference Image Quality Assessment Based On Phase Congruency And Local Entropies

With the development of internet and the popularization of mobile digital image displayer,such as smart phones,images have become an indispensible part of our daily lives for representing and transmitting information.However,subject assessment for image quality is very laborious,time consuming and impossible for real time assessment for large scale images,hence developing algorithms which can precisely,automatically and efficiently predicting the perceptive qualities if images are necessary.No-reference image quality assessment possesses great research value due to its practicability.This paper firstly simulate the algorithm SSEQ which is based on local special entropy and local spectral entropy,and test its performance,the experiment result meets our expectation.Then explore the phase congruency which can reflect the sensitively cognitive regions in images.Applying the phase congruency to no-reference image assessment,we propose a no-reference image quality assessment algorithm based on phase congruency and local entropies.The experiment result demonstrates that the proposed algorithm correlates well with human subjective scores.Finally,we studied the sparse representation model given that sparse representation is one of the most typical features of human visual systems.We proposed an opinion-unaware no-reference image quality assessment algorithm which utilizes quality aware features from three different domains and constructs the predicting model with sparse representation.The experiment result demonstrates that the proposed algorithm correlates well with subjective scores and has a good data base extensibility.