图像融合质量客观评价方法  

Objective quality assessment of image fusion

作　　者：高绍姝[1] 金伟其[1] 王岭雪[1] 王吉晖[1] 王霞[1] 

机构地区：[1]北京理工大学光电学院,光电成像技术与系统教育部重点实验室,北京100081

出　　处：《应用光学》2011年第4期671-677,共7页Journal of Applied Optics

基　　金：国家自然科学基金项目(60971010)

摘　　要：随着图像融合技术的快速发展,科学的图像融合质量客观评价对选择合理的融合算法以及研究新的融合算法具有重要的指导意义,成为图像质量评价研究的热点。理想的图像质量客观评价方法可以给出与人主观感受相一致的量化指标。对目前已提出的多种图像融合质量客观评价算法进行综述。简要介绍了基本的客观评价指标;将基于边缘保持度、基于结构相似度、基于信息论以及基于对比度的4类客观评价算法分别进行了介绍,着重分析和比较了各类算法的思路及特点;总结融合图像质量客观评价算法的研究趋势,指出无参考图像与彩色融合图像的质量评价方法是未来重要的发展方向。With the rapid development of image fusion technology,image fusion quality evaluation plays a very important guiding role in selecting or designing image fusion algorithms.Objective image quality assessment is an interesting research subject in the field of image quality assessment.The ideal objective evaluation method is consistent with human perceptual evaluation.The paper gives an overview of existing image fusion quality assessment algorithms.Firstly,basic objective evaluation specifications are presented briefly.Secondly,objective image quality assessment algorithms are classified into 4 categories： based on edge information preservation,based on structural similarity（SSIM）,based on information theory and based on contrast.They are introduced with emphasis on their strategies and characteristics.At last,the trends of future research are summarized.Objective image quality assessment considering features of the human visual system or based on specific visual tasks is more and more popular.Quality assessments of no reference image and color fusion image are important development directions in future.

关 键 词：图像融合 图像质量 客观评价 

分 类 号：TN911.73[电子电信—通信与信息系统]