基于尺度自适应CoF的红外与可见光图像融合算法研究

Research on infrared and visible light image fusion algorithm based on scale adaptive CoF

针对红外和可见光图像在传统多尺度域融合规则下容易损失图像边缘信息的问题,提出一种基于尺度自适应共现滤波器(Co-occurrence Filter,CoF)的非下采样剪切波变换(Non-subsampled Shearlet Transform,NSST)图像融合方法.这种尺度自适应CoF利用图像的归一化梯度调整滤波尺度,在保留原始CoF结构信息保持能力和去噪能力的基础上,保持密集的大梯度边缘信息.然后将其作为预处理步骤与NSST图像融合算法相结合,实现具有良好边缘保持性能的红外与可见光图像融合.实验表明,本文算法能有效保留源图像的边缘信息,并且通过对比实验分析验证了该算法的有效性.

Aiming at the problem that the fused images of infrared and visible light images under traditional multi-scale domain fusion rules are easy to lose image edge information,an adaptive scale Co-occurrence Filter(CoF)based non-subsampled shearlet transform(NSST)infrade-visible fusion algorithm is proposed.The prosed scale-adaptive CoF strategry adjusts the filter scale using the normalized gradient of the image,which not only retains the original CoF′s structure information retention and denoising ability,but also has the ability to maintain the dense large gradient edge information.The prosed scale-adaptive CoF strategry is then combined as a preprocessing step with the NSST image fusion algorithm in order to realize great edge-maintaining performance.Experiments indicate that the algorithm proposed can effectively preserve the edge information of the source image,and the effectiveness of the algorithm is verified by comparative experimentals.