Image quality assessment that aims to evaluate the image quality automatically by a computational model plays a significant role in image processing systems. To meet the need of accuracy and effectiveness, in the proposed method, complementary features including histogram of oriented gradient, edge information, and color information are employed for joint representation of the image quality. Afterwards, the dissimilarities of the extracted features between the distorted and reference images are quantified. Finally, support vector regression is used for distortion indices fusion and objective quality mapping. Experimental results validate that the proposed method outperforms the state-of-the-art methods in terms of consistency with subjective perception and robustness across various databases and different distortion types.

Image quality assessment is of fundamental importance for various image processing applications. A novel method is presented in which the joint occurrences of statistical local representation by log-Gabor filters and texture analysis by local tetra patterns and histograms of colour are considered as quality-aware features. Then the dissimilarities of these features between the distorted and reference images are quantified and mapped into quality score prediction by utilising a support vector regression. Extensive experiments on LIVE, CSIQ and TID databases show that the proposed method is remarkably consistent with human perception and outperforms many state-of-the-art methods, and also it is robust across different distortion types and different databases.