The Research Of Image Quality Assessment Method Based On Visual Saliency

In recent years, with the continuous development of digital image technology, people can quickly and easily get all kinds of images,videos and other visual information. Therefore, image quality evaluation in image acquisition, expression, storage and transmission of information processing has took an increasingly important part in the field of image processing. Many international and domestic experts and scholars made determined efforts to explore in-depth research on image quality evaluation method, and developed a lot of image quality evaluation models for many application area. Image quality evaluation has widely used in intelligent transportation, security, biomedical, military and other fields. Moreover, there are still some details can be further improved to achieve better performance.Based on the in-depth studies in the principles of human visual perception system, visual physiological, psychological physics,we restrict focus on adopting a more comprehensive indexes to evaluate the quality of images. According to analysis of several popular image quality evaluation models, especially SSIM( Structural Similarity), we have a good knowledge of their advantages, disadvantages and scope of application. Furthermore, we put forward a new similarity significant visual quality evaluation model based on the visual saliency, in which we replace contrast and brightness information with clarity and the average of brightness. In this article, we use the LIVE database to simulate the practical application. The results of IQAM(Image Quality Assessment Model) algorithm and subjective visual evaluation methods have good consistency, and the IQAM algorithm is valid to grayscale and color images. Experimental results demonstrate the effectiveness and efficiency of IQAM algorithm.