基于非下采样轮廓波变换和直觉模糊集的红外与可见光图像融合

Fusion of Infrared and Visible Images Based on Non-subsampled Contourlet Transform and Intuitionistic Fuzzy Set

针对传统图像融合方法造成的边缘模糊、细节损失、图像对比度与清晰度容易降低等问题,利用非下采样轮廓波变换,提出一种基于直觉模糊集和区域对比度的红外与可见光图像融合算法.首先,使用非下采样轮廓波变换将源图像分解,分别得到源图像的高频和低频成分.其次,利用直觉模糊集灵活准确描述模糊概念的特性,构建双高斯隶属函数对低频成分进行融合;利用区域对比度详细描述图像纹理信息的特点,采用多区域特征对比度结合距离分析的融合规则,对高频成分进行融合.最后使用非下采样轮廓波逆变换得到融合图像.实验结果表明,与其它融合算法相比,该算法提高了图像对比度,保留了源图像中的边缘和细节信息,且得到的融合结果具有更优的客观评价值.

Considering the traditional image fusion methods easily reduce the contrast and sharpness of image,blur edge and loss details,a fusion method of infrared and visible images is proposed based on intuitionistic fuzzy set and regional contrast in the fusion framework of non-subsampled contourlet transform. Firstly,the high and low frequency components of source images are obtained by using non-subsampled contourlet transform.Then the low frequency components are combined by intuitionistic fuzzy set including double-Gaussian function due to the characteristic that intuitionistic fuzzy set can describe the fuzzy concept flexibly and accurately. The high frequency components are combined by regional features contrast and Euclidean distance method because of the feature that the method can describe the image texture information in detail. Finally the fusion image is obtained by performing inverse non-subsampled contourlet transform. The experimental results show that the proposed fusion method outperforms traditional methods which deepens the contrast of images,retains the edge and detail information in the source image,and has a better evaluation value.