Research On Image Objective Quality Assessment

In recent years, with the rapid development of communication technology, the requirements for transmitted information are getting higher and higher. The image itself plays an important role in the transmission of information because it covers a large number of information. In addition, the image quality is also very significant for information dissemination. Therefore, the image quality assessment technology has very important significance for the wide application of image and video information.As a common one of the appraisal methods of image quality, the subjective assessment method obtain assessment results based on people's subjective feelings of the image. Although its accuracy and reliability are relatively good, it is time-consuming and laborious, which causes it not to be widely used. The objective assessment method is put forward for overcoming the shortages of the subjective method. The objective assessment method can assess the distorted image accurately through the establishment of mathematical model and the application of computer technology. This paper focuses on the research of full-reference and no-reference image quality assessments, and the main work of this paper includes:We describe the research background and significance of the image quality assessment, and analyze the research status. What's more, the subjective and objective assessment methods are briefly summarized.We analyze the full reference image quality assessment method based on structural similarity (SSIM). To eliminate the incomplete acquisition of the image structure information in SSIM algorithm, we propose the full reference image quality assessment method based on the structural similarity of gradient singular value decomposition (GSSSIM). In addition, we compare the SSIM and GSSSIM algorithms through simulation and fitting performance analysis. The results show that GSSSIM is more accurate than the image assessment results of SSIM in the LIVE Database Rlease2.We present the full reference image quality assessment method based on the structural similarity of the wavelet and singular value decomposition (WSSSIM). Through the simulation analysis in the LIVE Database Rlease2, it is proved that it can accurately assess the images of the JPEG2000 compression, FastFading Rayleigh channel distortion and Gauss white noise distortion.We put forward the no reference image quality assessment method based on the structural similarity of gradient singular value decomposition(SSSIM). The algorithm can obtain the accurate image quality assessment results in the Gauss distortion images. In addition, we make the simulation and performance analysis of the Gauss distortion images in the LIVE2 database and TID2013 database and prove the good performance of the algorithm.