基于双引导滤波的红外和可见光图像融合算法

图像融合是将多幅图像中有用或互补信息整合成一幅图像的过程。本文提出了一种基于引导滤波多尺度分解的红外和可见光图像融合算法。在传统的引导滤波图像融合算法的基础之上,利用双引导滤波器代替均值滤波器将源图像分解为小尺度纹理细节、大尺度边缘和基础图像；直接利用纹理细节及边缘层图像构建显著性映射图,用其代替额外的特征提取操作,可很好地突出源图像显著性信息的同时大大降低算法复杂度；利用显著性映射图、Sigmoid函数构造权重图,将源图像中具有视觉意义的信息注入到融合图像中；利用色彩模型转换融合方式,可更好保留图像的色彩信息。定性和定量实验结果证明,相比于传统的基于引导滤波的图像融合算法,本文算法的融合效果得到进一步提升。

Infrared and visible image fusion algorithmbased on dual guided filter

Image fusion is the process of integrating useful or complementary information from multiple images into one a single image.In this paper,an infrared and visible image fusion algorithm based on guided filter multi scale decomposition is proposed.Based on the traditional guided filter image fusion algorithm,a dual guided filter is utilized to replace the mean filter to decompose the source image into small scale texture details,large scale edges and basic images.The texture details and edge layer images are directly used to construct saliency maps,which can be instead of additional feature extraction operations to highlight the saliency information of the source image while greatly reducing the complexity of the algorithm.Using saliency map and sigmoid function to construct weight map,the visually meaningful information in the source image is injected into the fusion image.The color information of images can be better preserved by using the color model transformation and fusion method.Qualitative and quantitative experimental results show that the fusion effect of the proposed algorithm is further improved compared with the traditional guided filtering image fusion algorithm.