基于Contourlet变换和模糊理论的图像融合算法

Image fusion based on contourlet transform and fuzzy theory

提出一种基于Contourlet变换和模糊理论的红外与可见光图像融合算法。首先,对源图像进行Contourlet变换;随后对低频和高频的小波系数采用不同的融合规则,即低频部分采用基于模糊集的自适应加权融合方法,高频部分采用基于区域小波能量加权的方法;最后再通过contourlet逆变换,获得融合图像。实验结果表明该算法能够较好地保留源图像的细节信息,具有较好的视觉效果,是一种有效实用的方法。

For the fusion of visual and infrared images, a new method is proposed based on the contourlet transform and fuzzy theory. Firstly, the source images are decomposed using contourlet transform. Secondly, the contourlet coeffi- cients are fused with different rules. Fuzzy theory is used to fuse low frequency sub-band,which can adaptively calcu- late the wavelet coefficients weights of original images. While the high frequency sub-bands are merged with local en- ergy as weight. Ultimately, the fusion image is obtained through an inversed contourlet transform. The experimental re- suits show that the fusion scheme can get better visual effect and preserve image detail information of the original ima- ges. It is an effective and practical fusion method.