近自然彩色图像融合算法及其实时处理系统的发展  

Development of Near Natural Color Image Fusion Algorithms and Real-Time Image Fusion System

作　　者：倪国强[1] 肖蔓君[1] 秦庆旺[1] 黄光华[1] 

机构地区：[1]北京理工大学信息科学技术学院光电工程系,北京100081

出　　处：《光学学报》2007年第12期2101-2109,共9页Acta Optica Sinica

基　　金：国家自然科学基金(60377038)资助课题

摘　　要：在军事和生产领域,特别是在极其恶劣的可视条件下,图像融合技术越来越强调融合效果显示的"近自然"感觉,以及人类视觉感知的要求。从彩色图像融合算法的发展概况出发,着重介绍了两种具有代表性的基于视觉特性的近自然彩色图像融合方法,并特别阐述了课题组近期开展的相关算法研究成果。总结了国内外实时彩色图像融合系统的发展现状,概括了课题组已具备的硬件实验基础,综合分析了实时彩色图像融合技术的若干关键技术,如多源图像特征分析、实时图像配准、融合图像后处理等。针对实时图像融合技术的应用状况,提出图像融合系统未来的发展趋势。Along with the need in military and industry, especially in certain tough and dynamic environment, image fusion need emphases to display in near natural color results and consider the requirements of human visual perception. The development of color image fusion algorithms are introduced and two representative near natural color image fusion methods based on visual characters are described, and our researches on this field are summarized in detail. The progress on real-time color image fusion system and our hardware researches are specified. Several main technologies are discussed particularly, such as multi-source image characteristics analysis, real-time image registration, post processing of fused image and so on. With current application situation of image fusion, the prospect of image fusion technique and system are forecasted.

关 键 词：图像处理 图像融合 近自然彩色 视觉特性 图像配准 实时处理系统 

分 类 号：TP391[自动化与计算机技术—计算机应用技术] TN911.37[自动化与计算机技术—计算机科学与技术]