【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。更多还原

【Abstract】 Multi-scale analysis and compression sensing have become hot spots in recent years. and dictionary learning algorithm has been widely used in the field of image fusion. But when it is applied to the fusion of visible and infrared images,there often appear a lot of noise and edge ringing artifacts in their fusion results. Based on this,a fusion algorithm of infrared and visible light based on the Nonsubsampled Shearlet Transform( NSST) and dictionary learning is proposed in this paper. Firstly,the source images are decomposed with NSST; secondly,a rule based on regional energy is used in the low frequency sub-band coefficients after they block vectorization,zeroaveraging and sparse decomposition; and a rule,based on laplacian energy sum,high-frequency coefficients was fused; finally,the inverse NSST is used to get the final fused image. The simulation results show that the proposed algorithm is superior to other fusion algorithms in visual and objective evaluation.更多还原