Nature Scene Statistics Approach Based On ICA for No-Reference Image Quality Assessment

Abstract The no-reference/blind image quality assessment (NR-IQA) is the most difficult due to the reference images are not available. Nature scene statistics (NSS) has been proven successful in image modeling and feature extraction. However, classical NSS models could not capture the high-order dependencies reside in nature signals. In order to avoid this problem, we propose a NR-IQA algorithm with an independent component analysis (ICA) based NSS model. In the evaluations on LIVE database, experiment results show that the proposed approach outperforms stateof-the-art IQA algorithms.