基于非负矩阵分解和红外特征的图像融合方法

针对传统图像融合方法鲁棒性较差的缺点,提出了一种基于非负矩阵分解和红外特征的图像融合方法,实现源图像的目标区域和背景区域分别融合.首先将源图像作为原始数据进行非负矩阵分解得到特征基图像,特征基图像包含了源图像的整体特征;利用红外图像目标与背景灰度显著差异,通过区域生长方法从红外图像提取目标区域,将红外目标区域与特征基图像背景区域相融合得到融合图像.实验结果表明,该方法不仅简单易行,而且在保留了可见光的高空间分辨率和纹理细节信息同时,突出了红外图像对热目标敏感特点,提高了图像的可判读性.

A Image Fusion Method Based on Non-negative Matrix Factorization and Infrared Feature

A proposal method based on non-negative factorization(NMF) and infrared feature is presented for infrared and visual images fusion,which fuses the target region and background respectively.Source images are taken as source data in NMF analysis,from which the feature base image containing the global features of the original images can be extracted.Then the target regions from infrared image are segmented by region growing method.The segmented target regions are fused with background regions of feature base i...