基于人眼视觉系统的假彩色融合图像质量的评价方法

随着图像融合技术的发展,各种融合算法层出不穷,而很多情况下最终的融合图像是由人眼观察的,因此基于人眼视觉系统的图像融合质量评价显得尤为重要.为了能够模拟人眼对于融合图像的感知,得到融合后图像质量的客观评价,本文提出了一种基于色差理论的假彩色融合图像质量的评价方法.首先将源图像和融合图像转化到CIE L*a*b*均匀色空间,在频域对图像进行对比度敏感函数滤波,通过计算滤波后融合图像的色差判断图像的细节信息,在一定程度上色差越大信息越丰富；通过计算融合图像与源图像的色差判断融合图像与源图像的相关性,相关性越高,融合算法越好.通过融合图像的色差大小以及与源图像的相关性两个参量,得出融合算法的优劣.实验表明,与其他评价方法相比,本文提出的评价方法与人眼观察的结果较为一致.

A Novel Quality Evaluation Method for False Color Fused Image Based on Human Visual System

With the development of image fusion technology, various fusion methods are proposed. However, in many cases, the final fused image is observed by human eyes. As a result, fused image quality assessment based on human visual system is particularly important. In order to obtain the objective quality evaluation by simulating human eye perception, a novel method is proposed. First, the source images and the fused image are transformed into CIE L*a*b* color space respectively. Then in the frequency domain these images are filtered by the contrast sensitivity function. The richness of details of the fused image can be predicted by calculating the chromatism of it. With the increasement of the chromatism value, the information details also increase. While the correlation coefficient is evaluated at the same time by calculating the chromatism among the source images and the fused image. The higher the correlation coefficient, the better the fusion algorithm. With these two constrained parameters, a judgement can be given to different evaluation algorithms. Experiments show that the results of the proposed method are consistent with the human eye observation.