基于NSCT的自适应可见光与红外图像融合方法  

Self-adaption Fusion Method of Optical and Infrared Images Based on NSCT

作　　者：高永光 宋志娜[2] 蔡肖芋 

机构地区：[1]61683部队 [2]武汉大学遥感信息工程学院 [3]武汉大学测绘遥感信息工程国家重点实验室

出　　处：《地理空间信息》2016年第12期30-32,共3页Geospatial Information

基　　金：国家重点基础研究发展规划项目(2012CB719906);国家高技术研究发展计划资助项目(SS2013AA122301)

摘　　要：针对可见光与红外图像的融合问题,基于NSCT多方向、多尺度分解和平移不变性等优点,提出了一种基于NSCT的自适应可见光与红外图像的融合方法。通过NSCT对可见光与红外图像进行分解,在各分解层上根据图像质量评价指标来选取融合规则,并通过评价指标的自适应控制NSCT分解层数使最终融合图像具有最优的效果。实验结果表明,与传统融合算法相比,该方法在主观和客观评价方面均优于其他融合方法。Focusing on the image fusion problem of optical and infrared images,based on the advantages of multi-direction,multi-scale decomposition and translation invariance of NSCT,a self-adaption fusion method of optical and infrared images based on NSCT was presented in this paper.Firstly,optical and infrared images were decomposed by NSCT.And then,the fusion rules were selected on each decomposition level in line with the image quality evaluation.At last,we controlled the number of decomposition levels through quality evaluation to obtain the optimal fusion image.Experimental results show that the proposed method is superior to other fusion methods both in subjective and objective aspects.

关 键 词：红外图像 图像融合 NSCT 质量评价 

分 类 号：P237[天文地球—摄影测量与遥感]