Study On Image Quality Assessment Metrics Based On The Human Visual System

Image will produce different distortions in the process of image compression, transmission and processing, which affects the quality of information transmission. Therefore, it is important to study image quality assessment technology. Image subjective assessment method is the most reasonable and accurate and can really reflect visual image quality, but more time-consuming, costly and difficult to implement in practice. Image objective assessment method has the characteristics of speed, low cost, wide applications, assessment result reproducible and being unaffected by subjective factors. Subjective scores are a validation of image objective assessment method. Image quality objective assessment can be used to validate the performance and the associated algorithm for image processing system. Combining with the human visual characteristics, we building a corresponding visual model to further analyze and study image objective quality assessment algorithms to ensure that they are consistent with human perceptual quality assessment, more persuasive. For studying the human visual system(HVS), these have an important theoretical significance. Under such a premise, the paper carries out image quality assessment algorithms based on HVS.Firstly, the background and significance of studying image quality assessment is presented together with image quality assessment based on HVS, reviewing the status of image quality assessment algorithms. This paper introduces the study on the basis of these theories such as image and its quality, the structure of the human visual system and visual mechanism feature, public image quality database and the performance metrics of image quality assessment algorithms, exploring image artifacts after wireless transmission related to the proposed algorithm. Then, the classical image quality assessment algorithms are studied and made a performance comparison in the public image database. This paper gives an analysis to the performances of these algorithms as the theoretical basis for the proposed algorithm.Secondly, in order to improve image information uncertainty measurement of the Multiple-scale Structural SIMilarity(MSSIM), a novel algorithm called i MSSIM based on internal generative mechanism is proposed, combining with Human Visual System(HVS). Internal generative mechanism based on the Piecewise Auto Regressive(PAR) model decomposes distorted image and the original image into two parts, the predicted part of image content using MSSIM algorithm assessment and image information uncertainty Part using PSNR assessment. Then, Mean Square Error is used as weight to combine the two scores to acquire the overall image quality assessment results. Using different performance index verifies the consistency of objective and subjective assessment. Then, the proposed algorithm is on the public image database for simulation and makes a performance comparison with the existing those, demonstrating the performance of the improved algorithm has been further improved.Finally, in order to improve human eye insensitive part measurement of the Visual Information Fidelity(VIF), a novel algorithm called i VIF based on Internal Generative Mechanism is proposed, combining with effectiveness of the image assessment after wireless transmission and Human Visual System(HVS) and studying the image distortions because of wireless transmission. Experiments on the Wireless Image Quality(WIQ) database demonstrate that the improved algorithm is superior to the existing those.