一种NSCT域多聚焦图像融合新方法

针对多聚焦图像融合存在的问题,提出一种基于非下采样Contourlet变换（NSCT）的多聚焦图像融合新方法。首先,采用NSCT对多聚焦图像进行分解;然后,对低频系数采用基于改进拉普拉斯能量和（SML）的视觉特征对比度进行融合,对高频系数采用基于二维Log-Gabor能量进行融合;最后,对得到的融合系数进行重构得到融合图像。实验结果表明,无论是运用视觉的主观评价,还是基于互信息、边缘信息保留值等客观评价标准,该文所提方法都优于传统的离散小波变换、平移不变离散小波变换、NSCT等融合方法。

A Novel NSCT-Based Technique for Multi-Focus Image Fusion

In order to overcome the shortcoming of the traditional multi-focus image fusion methods, a novel non-subsampled contourlet transform（NSCT） based technique for multi-focus image fusion is proposed. Firstly, the source multi-focus images are decomposed by using the NSCT. Secondly, the low-frequency subband coefficients are fused by the local visual contrast mechanism based Sum-Modified-Laplacian（SML）, and the high-frequency subband coefficients are fused by a 2D Log-Gabor energy rule. Finally, the inverse NSCT is employed to reconstruct the fused image. Experimental results demonstrate that the proposed method is better than series of popularly used fusion methods, including discrete wavelet transform, shift invariant DWT, NSCT, etc, in terms of both subjective and objective evaluations.