红外与可见光图像的非采样Contourlet变换融合方法  

Infrared and Visible Images Fusion Method Based on the Nonsubsampled Contourlet Transform

作　　者：陈优良[1] 

机构地区：[1]国网江西省电力公司拓林水电厂,江西永修330317

出　　处：《南昌航空大学学报（自然科学版）》2015年第2期33-38,共6页Journal of Nanchang Hangkong University(Natural Sciences)

摘　　要：针对传统基于非采样Contourlet变换图像融合技术存在的问题,提出改进的基于图像局部能量的非采样Contourlet变换红外与可见光图像融合方法。首先对待融合图像进行非采样Contourlet变换;然后在图像低频成分采用基于图像局部能量的加权平均融合准则进行图像融合,在图像高频成分采用归一化将高频系数的能量差与设定的融合阈值进行比较,再进行分步融合;最后进行非采样Contourlet变换的重构得出融合图像。实验章节通过合成及真实图像算例证明本文方法相对于传统的图像融合方法具有更好的融合效果及精度。According to the disadvantage of the traditional fusion method, It was proposed an improved visible and infrared images fusion method based on the Nonsubsampled Contourlet Transform. Firstly, we decomposed the images to be fused with the Nonsubsampled Contourlet Transform, and then we implemented the image fusion by doing the weighted average based on the local energy in the Low Frequency part of the images and doing the stepwise fusion based on the normalization threshold comparison. Finally, we obtained the fusion image by reconstructing the Nonsubsampled Contourlet Transform. The synthetic and real images experiments showed the proposed method has the better fusion effect and accuracy than the traditional image fusion methods.

关 键 词：红外图像 可将光图像 非采样CONTOURLET变换 图像融合 

分 类 号：TN911.73[电子电信—通信与信息系统]