基于FFST和方向特性对比度的图像融合算法

为了克服红外可见光图像融合方法存在的不足,结合快速有限剪切波变换(Fast Finite Shearlet Transform,FFST)的平移不变性以及较高的方向敏感性,提出了一种基于快速有限剪切波变换域的自适应多方向图像融合新方法。首先,对严格配准后的图像进行快速有限剪切波变换分解,得到低频子带和高频子带系数；然后,对低频子带系数采用非负矩阵分解的一个约束稀疏算法,即在基本非负矩阵分解的优化函数中施加稀疏性约束,使分解更优,以此来提高重构后图像的清晰度；高频子带系数则采用联合方向特性的对比度进行选取,以得到丰富的细节信息。最后,利用快速有限剪切波逆变换得到重构后的图像。实验结果表明,融合后的图像充分结合了源图像的有用信息,整体轮廓清晰,在客观评价上也有一定的提高。

Fusion algorithm based on FFST and directional characteristics contrast

To overcome the shortcomings of the infrared and visible image fusion methods,combining with shift-invariance and higher directional sensitivity of fast finite shearlet transform(FFST),a novel adaptive multi-directional image fusion method based on FFST domain was proposed.Firstly,the FFST is used to decompose the strict registered images,and the low frequency sub-band coefficients and the high sub-band coefficients are obtained.Secondly,the fusion principle of low frequency sub-band coefficients is based on a constrained sparse algorithm for nonnegative matrix factorization,that is,a sparseness constraint is added to the original nonnegative matrix factorization algorithm,making the decomposition better,which can improve the definition of the image reconstruction.In order to obtain the detail information,a combined directional characteristics contrast is adopted to choose high frequency sub-band coefficients. Finally,the fused image is reconstructed by inverse FFST.The results show that the fused image reserves useful information of the source image and has clear overall outline.