基于人眼视觉特性的融合图像质量评价算法

摘要

针对彩色融合图像在人眼视觉特性上的表现,分析现有图像融合质量评价方法特点,提出一种基于对比度敏感度特性加权结构相似度的彩色融合图像质量评价算法.该方法将图像颜色的亮度分量和色度分量分开,首先对图像亮度分量的结构相似度进行对比度敏感特性加权,再与色度分量的结构相似度进行加权以获得图像的结构相似度.该方法充分考虑人眼视觉特性和颜色特性,对不同图像融合评价方法的评价结果表明,该方法能较好地完成彩色融合图像的质量评价.%In allusion to the performance of the color fusion image on the human visual features, the characteristic of the existing evaluation method on image fusion is analyzed. An evaluation algorithm for color fusion image is proposed based on contrast sensitivity characteristic weighted structure similarity. The brightness component and the chromaticity component of image color are separated. The structure similarity of the image brightness component is first weighted by the contrast sensitivity characteristic, and then it is weighted with the structure similarity of the chromaticity component to acquire the structure similarity of the image. The experiments of different evaluation methods for image fusion demonstrate that this method can accomplish the quality evaluation of the color fusion image preferably.