## 基于联合稀疏表示的红外与可见光图像融合研究

Title  Fusion Research of Infrared and Visible Images Based on Joint Sparse Representation
Abstract Fusion of infrared and visible images is an important branch of image fusion field, which has wide applications in  military and civilian fields such as aviation, remote sensing and so on. To find a suitable fusion algorithm for infrared and visible images,  set out  from the images’ joint sparsity, the image fusion algorithm will be further research in this paper. Based  on the system study on basic principle of the image fusion, combined with the sparse representation theory which has superior performance in current,  this paper focuses  on a joint  sparse representation-based image fusion method for   infrared and visible images. Firstly, train the over-completed dictionary by K-SVD algorithm, then solve the sparse coefficients with optimization algorithm, OMP, the coefficients include common sparse coefficient and unique sparse coefficients, then use the weighted average rule fuse  unique factors, finally, reconstruct the fuse image by fusion coefficient and over-complete dictionary. Experimental result shows that the fusion method in this paper can obtain a fusion image with the prominent infrared target and clear background details,  and both with  the superior denoise function.

Keywords   Fusion of Infrared and Visible Images, Joint Sparse Representation, Common and Innovation Sparse coefficients, K-SVD, OMP

1绪论1
1.1研究背景及意义.1
1.2国内外研究现状.2
1.3本文主要工作及章节安排.3
2图像融合基本原理5
2.1融合层次.5
2.2融合算法.5
2.2.1基于小波变换的图像融合算法6
2.2.2基于ICA的图像融合算法6
2.2.3基于稀疏表示的图像融合算法7
2.3融合规则7
2.3.1基于单像素点的融合规则.7
2.3.2基于区域的融合规则.8
2.4图像去噪9
2.4.1空间域去噪9
2.4.2变换域去噪.10
2.5评价体系.11
2.5.1主观评价指标11
2.5.2客观评价指标11
2.6本章小结.14
3稀疏表示理论15
3.1稀疏模型15
3.2优化算法15
3.2.1贪婪追踪算法.15
3.2.2 基于联合稀疏表示的红外与可见光图像融合研究:http://www.lwfree.cn/tongxin/20181202/26884.html
------分隔线----------------------------