基于FRGF和改进PCNN的红外可见光图像融合

为解决图像融合中边缘细节保留不理想的问题,本文提出了一种快速滚动引导滤波器和改进脉冲耦合神经网络相结合的红外可见光图像融合方法。提出的快速滚动引导滤波器可以较好地在保留边缘、细节纹理信息的同时有效提高运行效率。首先,利用快速滚动引导滤波和高斯滤波对源图像进行多尺度分解；其次,为了使基础层图像更好地突出轮廓信息,采用相似性匹配的融合规则对图像进行融合；然后,细节层采用改进参数自适应脉冲耦合神经网络规则进行融合；最后,经过多尺度重构得到融合结果图。实验结果表明,与其它5种融合方法相比,该算法不仅在视觉效果上得到了提升,而且能够充分保存图像的边缘和纹理等信息,极大地提高了运行效率。另外,该方法在客观评价指标上均优于对比方法。

Infrared and visible image fusion based onFRGF and improved PCNN

In order to solve the problem that edge details are not ideally preserved in image fusion,an infrared and visible image fusion method based on fast rolling guided filter and improved pulse coupled neural network is proposed in this paper.The proposed fast rolling guided filter can better retain the edge and detail texture information while effectively improving operating efficiency.Firstly,the multi scale decomposition of the source images is carried out by fast rolling guided filter and Gaussian filter.Secondly,in order to make the base layer image better highlight the contour information,the fusion rule of similarity matching is used to fuse the images,and then the detail layers are fused with the improved parameter adaptive pulse coupled neural network rule.Finally,the fusion result is obtained by multi scale reconstruction.The experimental results show that compared with other five fusion methods,the proposed method not only improves the visual effects,but also fully saves information such as the edges and textures of the image,and greatly improves the running efficiency.In addition,the method on the objective evaluation indexes is superior to the comparison method.