NSST域模糊逻辑的红外与可见光图像融合

Infrared and visible light images fusion of fuzzy logic on NSST domain

为了在红外与可见光图像融合时保留各自更多的细节信息,同时降低算法复杂度,采用了非下采样剪切波变换(NSST)和改进模糊逻辑的红外与可见光图像融合方法,利用NSST算法对红外图像和可见光图像分别进行多尺度、多方向稀疏分解,分别得到低频子带系数和高频子带系数。然后对低频子带系数采用基于改进的模糊柯西隶属函数的权值平均融合规则;对高频子带系数采用能量匹配度和视觉敏感度系数相结合的融合规则。最后对低频子带融合系数和高频子带融合系数执行NSST逆变换得到最终的融合图像,并进行了理论分析和实验验证。结果表明,此融合方法不仅可以保证融合清晰度,对缩短算法的运行时间也是有帮助的。

In order to retain more detail information and reduce the algorithm complexity when fusing the infrared and visible light images, a fusion algorithm based on non-subsampled shearlet transform （ NSST ） and improved fuzzy logic was proposed to decompose source images sparsely on multi-direction and multi-scale. Low-frequency subband coeffients and high-frequency subband coeffients were obtained. The improved average fusion method of fuzzy Cauchy membership function was adopted in low-frequency subband coeffients. The fusion rule of the combination of energy compatibility and visual sensitivity coefficient was used in high-frequency subband coeffients. Finally, fusion image was obtained after NSST inverse transformation. Experimental results show that the fusion method can not only guarantee the definition of fused image, but also shorten the running time of algorithm.