一种基于显著性的多尺度图像融合模型  

A Saliency-Based Multi-scale Image Fusion Model

作　　者：李蕴奇[1] 

机构地区：[1]吉林省经济信息中心,长春130061

出　　处：《吉林大学学报（理学版）》2013年第2期285-288,共4页Journal of Jilin University:Science Edition

基　　金：国家自然科学基金(批准号:40972059)

摘　　要：提出一种基于显著性的多尺度图像融合模型.先利用轮廓波变换将输入图像分解成子图序列,然后计算每幅子图像的区域显著性,并选取显著性大的参数作为最终融合参数,最后通过反变换获得融合图像,图像的显著性通过谱冗余法获得.实验结果表明,该方法较传统方法融合效果更好.This paper presents a multi-scale saliency-based image fusion model.First,the source images were decomposed into sub-image sets by contourlet transform,then the saliency of each piece of sub-image was calculated and the coefficients were selected by maximizing saliency,and finally,the fused image was obtained by inversing contourlet transform.The saliency of the image was calculated by spectral residual theory.Experimental results show that the proposed model outperforms traditional models.

关 键 词：图像融合 显著性 轮廓波变换 谱冗余 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]