摘要
针对现有的红外与可见光图像融合方法存在源图像细节纹理丢失、对比度和清晰度不高等问题,提出一种基于鲁棒主成分分析(Robust Principal Component Analysis,RPCA)和参数化对数图像处理(Parameterized Logarithmic Image Processing,PLIP)的红外与可见光图像融合方法。利用RPCA将待融合图像分解为稀疏矩阵和低秩矩阵。对稀疏矩阵结合局部二值模式(Local Binary Pattern,LBP)和直觉模糊集(Intuitionistic Fuzzy Set,IFS)进行融合,以减少噪声干扰,获取丰富的细节。对低秩图像基于主成分分析(Principal Component Analysis,PCA)和PLIP融合,以增强融合图像背景的纹理信息。实验结果表明,所提方法在保留待融合图像的边缘和细节、融合图像的对比度和清晰度等方面均有所提高,并且主观感受更自然。
The existing infrared and visible image fusion algorithms have the problems of loss of source image details texture,poor contrast and sharpness.An infrared and visible image fusion algorithm is proposed based on the combination of the robust principal component analysis(RPCA)and the parameterized logarithmic image processing(PLIP).The source images were decomposed by the RPCA to obtain the corresponding sparse and low-rank matrices.The sparse matrices were fused with LBP and IFS to reduce noise interference and obtain rich details.The low rank images were fused based on the principal component analysis(PCA)and the PLIP to enhance the texture information of the fused image background.Experimental results show that the proposed algorithm is superior to the comparison algorithms in preserving the contour and texture details of the source images,improving the sharpness and contrast of the fused images.The subjective perception is more natural.
作者
许静
刘敬
李康欣
XU Jing;LIU Jing;LI Kangxin(School of Electronic Engineering,Xi’an University of Posts and Telecommunications,Xi’an 710121,China)
出处
《西安邮电大学学报》
2022年第6期22-30,共9页
Journal of Xi’an University of Posts and Telecommunications
基金
陕西省自然科学基础研究计划项目(2021JM-459)。
关键词
图像融合
红外与可见光图像
鲁棒主成分分析
参数化对数图像处理
image fusion
infrared and visible images
robust principal component analysis
parameterized logarithmic image processing
作者简介
许静(1977-),女,硕士,讲师,从事数字图像处理研究。E-mail:xujing@xupt.edu.cn;刘敬(1975-),女,博士,副教授,从事遥感图像处理研究。E-mail:zyhalj1975@163.com。
