Natural image quality assessment based on visual cognitive mechanism

In view of the main problems existed at present in no-reference (NR) natural image quality assessment (IQA), This paper proposes a more general-purpose, efficient and integrated resolution based on visual cognitive mechanism. Firstly, it puts forward a inspiring visual cognitive computing model (IVCCM) based on visual heuristic principles. Secondly, it presents a asymmetric generalized Gaussian mixture distribution model (AGGMD) for natural images. Thirdly, it extracts quality-aware features from natural images and form Quality-aware Uniform Features Descriptors (QAUFD). Fourthly, it realizes the resolution with IVCCM, AGGDM and QAUFD to NR IQA. Experimental results show that the proposed resolution correlates better with human perceptual measures.