一种基于NSST和字典学习的红外和可见光图像融合算法
发布时间：2019-01-15 20:45
【摘要】：近年来随着多尺度分析和压缩感知成为研究的热点,字典学习算法在图像融合领域得到了广泛应用,但是其算法应用于可见光和红外图像的融合,容易出现块状噪声,边缘有振铃现象。基于此,本文提出了一种基于非下采样剪切波变换(NSST)和字典学习的红外和见光图像融合算法研究,对NSST分解的低频分量利用滑动窗口得到图像块序列,并对其进行零均值化后再稀疏分解,选择区域能量的融合规则,高频子带选择拉普拉斯能量和的融合规则。仿真结果表明,本文的算法在视觉和客观评价指标上优于现有几种融合算法。
[Abstract]:In recent years, with the research of multi-scale analysis and compression perception, dictionary learning algorithm has been widely used in the field of image fusion, but its algorithm is applied to the fusion of visible and infrared images, which is prone to appear block noise. There is a ringing on the edge. Based on this, this paper proposes an infrared and visible image fusion algorithm based on non-downsampling shear wave transform (NSST) and dictionary learning. The low frequency components of NSST decomposition are obtained by sliding window. After zero mean decomposition, the fusion rules of region energy and Laplacian energy sum are selected in the high frequency subband and the fusion rule of Laplace energy sum is chosen. The simulation results show that the proposed algorithm is superior to the existing fusion algorithms in visual and objective evaluation.
【作者单位】： 西北工业大学电子信息学院;西北工业大学保密处;西北工业大学计算机学院;
【基金】：国家自然科学基金(61071171)资助
【分类号】：TP391.41
[Abstract]:In recent years, with the research of multi-scale analysis and compression perception, dictionary learning algorithm has been widely used in the field of image fusion, but its algorithm is applied to the fusion of visible and infrared images, which is prone to appear block noise. There is a ringing on the edge. Based on this, this paper proposes an infrared and visible image fusion algorithm based on non-downsampling shear wave transform (NSST) and dictionary learning. The low frequency components of NSST decomposition are obtained by sliding window. After zero mean decomposition, the fusion rules of region energy and Laplacian energy sum are selected in the high frequency subband and the fusion rule of Laplace energy sum is chosen. The simulation results show that the proposed algorithm is superior to the existing fusion algorithms in visual and objective evaluation.
【作者单位】： 西北工业大学电子信息学院;西北工业大学保密处;西北工业大学计算机学院;
【基金】：国家自然科学基金(61071171)资助
【分类号】：TP391.41
【参考文献】
相关期刊论文 前5条
1 杨森林;万国宾;;基于分块压缩感知的遥感图像多尺度融合[J];西北大学学报(自然科学版);2015年03期
2 李晖晖;曾艳;杨宁;姚西文;钱林弘;;改进的压缩感知重构算法及其在图像融合中的应用[J];南京理工大学学报;2014年02期
3 王s，
本文编号：2409090
本文链接：https://www.wllwen.com/kejilunwen/ruanjiangongchenglunwen/2409090.html
