Image quality assessment via spatial structural analysis. (August 2018)

• Record Type: Journal Article
• Authors: Yang, Xichen; Sun, Quansen; Wang, Tianshu
• Is Part Of: Computers & electrical engineering. Volume 70(2018)
• Page Start: 349
• Page End: 365
• Publication Date: 2018-08
• Subjects: Full-reference -- Image quality assessment -- Distortion -- Spatial structural distribution -- Structural similarity; Computer engineering -- Periodicals; Electrical engineering -- Periodicals; Electrical engineering -- Data processing -- Periodicals; Ordinateurs -- Conception et construction -- Périodiques; Électrotechnique -- Périodiques; Électrotechnique -- Informatique -- Périodiques; Computer engineering; Electrical engineering; Electrical engineering -- Data processing; Periodicals; Electronic journals; 621.302854
• DOI: 10.1016/j.compeleceng.2016.08.014
• Languages: English
• ISSNs: 0045-7906

Abstract

Highlights: The proposed image quality assessment method takes both spatial contrast and structural distributions into consideration. Image gray-scale fluctuation primitive is introduced to analyze image gray-scale fluctuations, and the gray-scale fluctuation primitive map (GFM) of each image is obtained. The spatial structural information variation matrix is designed to measure the variation of image structural degradations caused by distortions. The proposed method is robust to the change of training set. The proposed method has a high correlation with human subjective judgments of diversely distorted images. Graphical abstract: Abstract: The human visual system is sensitive to image structural information. Modeling of image structural similarity has been regarded as suitable for achieving perceptual quality predictions. However, most structural similarity-based image quality assessment (IQA) methods focus on spatial contrast without fully considering the spatial structural distribution. Hence, we propose an IQA method that considers both spatial contrast and structural distributions. First, the image gray-scale fluctuation map (GFM) is calculated. Second, the spatial structural information variation matrices (SSVMs) between the GFMs of distorted and pristine images are obtained. Finally, the quality prediction model is trained using support vector regression (SVR). The experimental results show that the proposed method can accurately predict human perceptual image quality. Experiments on the LIVE2 database show that the Spearman rank-order correlation coefficient (SROCC) and linear correlation coefficient (LCC) values exceed 0.85, while the scale or distortion type of the training set changes, which indicates stability.