基于联合低秩稀疏分解的红外与可见光图像融合

为进一步提高红外与可见光融合图像的细节信息和整体对比度，降低伪影和噪声，考虑了红外与可见光图像的相关性，提出了一种基于联合低秩稀疏分解的红外与可见光图像融合方法。首先，利用联合低秩稀疏分解方法将红外和可见光源图像分别分解成共同低秩分量、特有低秩分量和特有稀疏分量；其次，利用非下采样Shearlet变换方法对特有低秩分量进行融合；然后，采用区域能量融合策略实现特有稀疏分量融合；最后，共有低秩分量与融合后的特有低秩分量和特有稀疏分量相加得到最终融合图像。在Nato-camp、Bristol Eden Project和TNO公共测试数据集上进行的实验测试了所提算法性能。实验结果表明，与其他9种融合方法相比，所提方法能够有效地提取红外图像中的目标信息和保留可见光图像的背景信息，熵、互信息、标准差、视觉信息保真度、差异相关系数之和和 Qy 客观评价指标明显优于对比方法。 

Infrared and Visible Image Fusion via Joint Low-rank and Sparse Decomposition

In order to further improve the detail information and overall contrast of the fused images and reduce artifacts and noises, an infrared and visible image fusion method based on joint low-rank and sparse decomposition was proposed by considering the correlation between infrared and visible images. First, infrared and visible images are jointly decomposed into common low-rank component, specific low-rank components and specific sparse components by using the joint low-rank and sparse decomposition method. Second, the nonsubsampled shearlet transform-based fusion method is performed on the specific low-rank components. Third, the specific sparse components are fused by adopting regional energy fusion rule. Finally, the fused image is obtained by integrating the common low-rank component, the fused specific low-rank component and the fused specific sparse component. The experiments conducted on the Nato-camp、Bristol Eden Project and TNO publicly test data sets are used to test the performance of the proposed algorithm. The experimental results demonstrate that the proposed method can effectively extract the target information of infrared image and retain the background of visible image compared with other nine fusion methods. Meanwhile, the values of the objective evaluation metrics such as entropy, mutual information, standard deviation, visual information fidelity, the sum of the correlations of differences and Qy are obviously better than those of the comparison methods.