基于视觉感知的最好质量图像评价方法

Evaluation Method of the Best Image Quality Based on Visional Perception

基于人类视觉系统对图像感知功能的研究,提出影响图像质量的4个参数。综合这4个参数在图像质量评价领域的作用,建立最好质量图像评价的数学方法。分析CAF值与4个参数之间的函数关系,并通过大量实验,认为CAF最大值所对应的图像就是最好的图像,而此结论与视觉感知是一致的。该方法不依赖参考图像,同时适用于灰度图和彩色图,方法简单、直观。

The paper is mainly based on the human visual system to image the function of perception （the four parameters of the image quality for short） put forward the image quality of a decision four quality parameters. Combination of four parameters to establish the best image quality evaluation of mathematical methods （CAF）, analyzing between the CAF value and the relationship with the four parameters and doing a lot of trials ultimately get the maximum CAF of image is the best one, and this conclusion is consistent with the visual perception. This method has the characteristics do not only rely on reference image is suitable for grayscale, but also suitable for color. The method is simple, intuitive, and innovative.