基于图像清晰度和结构相似度的融合图像质量评价.doc

文档介绍：
基于图像清晰度和结构相似度的融合图像
质量评价#
肖芬1,夏静2*
5
10
15
20
25
(1. 智能计算与信息处理教育部重点实验室(湘潭大学),湖南湘潭 411105;
2. 湘潭大学信息工程学院,湖南湘潭 411105)
摘要:在研究已有图像融合质量评价方法的基础上,提出一种基于清晰度和结构相似度的融
合图像质量评价方法。充分考虑图像结构信息和人眼视觉特性,通过分别计算图像块的清晰
度和协方差,对融合图像块的结构信息进行加权,从而得到融合图像的质量评价。该方法无
需理想源图像,通过对不同融合算法得到的融合图像进行质量评价,结果表明该方法是有效
的质量评价方法,能为融合算法选择提供可靠依据。
关键词:图像融合;图像质量评价;结构相似度;清晰度
中图分类号:TP37
Image Fusion Quality Assessment Based On Clarity And
Structural Similarity
Xiao Fen1, Xia Jing2
(1. Key Laboratory of puting & Information Processing of Ministry of Education,
Xiangtan University, HuNan XiangTan 411105;
2. College of Information Engineering, Xiangtan University, , HuNan XiangTan 411105)
Abstract: Based on the image clarity and structural similarity(SSIM), a new quality assessment for
fused image is proposed in this paper. The clarity and covariance of the image block is used for
assessment by taking into account human visual system characteristics. An then, SSIM scores are
weighted by the clarity and covariance for fused image quality assessment, furthermore the ideal
source image are not required parison. We demonstrate the proposed method by
incorporating it for fusion image which obtained from some classical fusion methods. The
experimental results show that our new approach provides results are highly consistent with
human subjective judgement of the fused images quality.
Keywords: image