基于前向-后向自校正扩散引导特征重建的图像融合  

An image fusion method using forward-backward self-correcting diffusion guided feature reconstruction

作　　者：张相博 刘刚 肖刚[2] ZHANG Xiang-bo;LIU Gang;XIAO Gang(College of Automation Engineering,Shanghai University of Electric Power,Shanghai 200090,China;School of Aeronautics and Astronautics,Shanghai Jiao Tong University,Shanghai 200240,China)

机构地区：[1]上海电力大学自动化工程学院,上海200090 [2]上海交通大学航空航天学院,上海200240

出　　处：《控制与决策》2022年第8期2134-2140,共7页Control and Decision

基　　金：国家自然科学基金项目(61673270,61203224);国家973计划项目(2014CB744903);上海浦江人才计划项目(16PJD028)。

摘　　要：现有图像融合方法不同程度地存在边缘阶梯效应,导致一些空间伪影引入融合图像.鉴于此,提出一种新的解决图像融合过程中鲁棒性差的方法—–前向-后向自校正扩散引导特征重建(forward-backward self-correcting diffusion,FBSD),对分解后各特征之间的差异设计一种基于期望值最大算法和主成分分析的混杂融合策略.最后利用评价指标评估所提出算法的性能,验证了所提出方法在边缘阶梯效应的处理上优于现有的图像融合方法,同时验证了融合决策的有效性.The existing image fusion methods have an edge ladder effect in varying degrees,which leads to some spatial artifacts introduced into the fused image.This paper proposes a new method to solve the poor robustness in the process of image fusion,that is,forward-backward self-correcting diffusion(FBSD) guided feature reconstruction.According to the differences between features after decomposition,a hybrid fusion strategy based on expectation-maximization algorithm and principal component analysis algorithm is designed.Finally,the evaluation index is used to evaluate the performance of the proposed algorithm,and it is verified that this method is better than the existing image fusion methods in dealing with the edge step effect,and the fusion decision is effective.

关 键 词：图像融合 特征重建 期望值最大算法 主成分分析 红外与可见光图像融合 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]