摘要
针对利用传统方法进行红外与可见光图像融合时易出现边缘模糊以及细节分辨能力弱的问题,提出一种基于小波变换和各向异性扩散的红外和可见光图像融合算法。首先,将红外图像和可见光图像利用小波变换进行多尺度分解,获取原图像所对应的高频部分和低频部分;其次,将高频部分和低频部分分别进行各向异性扩散,生成对应图像的基础层和细节层;然后,采用KL变换对异源图像的细节层进行融合,采用加权平均方法对基础层进行融合;最后,将融合后的细节层和基础层通过线性重构得到最终的融合图像。为了验证所提出算法的优势,将其与3种经典融合算法进行比较。通过大量融合实验表明,相比于其他3种经典融合算法,所提出的算法不仅实时性好,而且融合结果能够较好保留原图像丰富的细节信息,具有较高的清晰度。
In order to solve the problems of edge blur and weak detail resolution when using traditional methods to fuse infrared and visible images,an infrared and visible image fusion algorithm is proposed based on wavelet transform and anisotropic diffusion.Firstly,the wavelet transform is used to perform a multiscale decomposition on infrared image and visible image,and the corresponding high frequency and low frequency parts of the source images are obtained.Secondly,the high frequency part and the low frequency part are respectively anisotropic diffused to generate the basic layer and detail layer of source images.Then,KL transforming is used to fuse the detail layer of the heterogeneous image,and the weighted average method is used to fuse the base layer.Finally,the fusion image is generated by linear reconstruction of the fused detail layer and the base layer.A comparison between the proposed algorithm and three traditional ones has been made to confirm the advantages.A large number of fusion experiments show that the proposed algorithm is better in real-time and the fusion result can better reserve the very details of the source with higher clarity.
作者
郝帅
安倍逸
付周兴
马瑞泽
赵新生
马旭
刘彬
HAO Shuai;AN Beiyi;FU Zhouxing;MA Ruize;ZHAO Xinsheng;MA Xu;LIU Bin(College of Electrical and Control Engineering,Xi’an University of Science and Technology,Xi’an 710054,China;Xi’an Satellite Control Center,Xi’an 710043,China)
出处
《西安科技大学学报》
CAS
北大核心
2022年第1期184-190,共7页
Journal of Xi’an University of Science and Technology
基金
国家自然科学基金项目(51804250)
中国博士后科学基金项目(2019M653874XB,2020M683522)
陕西省科技计划项目(2020JQ-757,2021JQ-572)
陕西省教育厅科研计划项目(18JK0512)
陕西省创新能力支撑计划(2020TD-021)
西安市碑林区科技计划项目(GX2116)。
关键词
图像融合
小波变换
各向异性扩散
多尺度分解
image fusion
wavelet transform
anisotropic diffusion
multi-scale decomposition
