基于2D-PWVD的多光谱图像与全色图像融合  

2D-PWVD based method for multi-spectral image and panchromatic image fusion

作　　者：侯若楠 薛斌党[1] 

机构地区：[1]北京航空航天大学图像中心,北京100191

出　　处：《中国体视学与图像分析》2015年第4期327-338,共12页Chinese Journal of Stereology and Image Analysis

基　　金：航天支撑基金(No.20141502);中航创新基金(No.CXY2012BH02)

摘　　要：二维Wigner-Ville分布具有良好的空频特性,能够有效地对图像进行多频谱分析。为了更好地融合多光谱图像和全色图像,本文提出一种基于2D-PWVD的图像融合方法。首先对多光谱图像进行IHS变换,全色图像则转换为灰度图像;然后对多光谱图像I分量与全色图像分别进行2D-PWVD变换得到其频谱图;接着,对不同频率的频谱图按照融合准则进行融合得到新的频谱图,进而进行图像重构;最后将重构图像进行IHS反变换得到融合结果。融合实验结果表明,该方法既可以保留全色图像的分辨率,同时较好地保留了多光谱图像的光谱信息。Two-dimensional pseudo Wigner-Ville distribution （2D-PWVD） has good spatial-frequency properties, which is effective in analyzing the spectrum of images. In this paper, we present an image fu- sion method based on 2D-PWVD to integrate the advantages of the multi-spectral image and panchromatic image. Firstly IHS transform is performed on the multi-spectral image to achieve its intensity （I） compo- nent, while the panchromatic image is transformed to gray image. Then we get the spectrum diagram of the component I and panchromatic image respectively through 2D-PWVD transformation. A new spectrum dia- gram is achieved according to the fusion rule of different frequencies, and then the image is reconstructed Finally, the fused image is obtained by inverse IHS transform of the constructed image. Experiment results show that the fused images can not only retain the resolution of panchromatic image, but also preserve the spectral information of multi-spectral image.

关 键 词：多光谱图像 全色图像 图像融合 2D-PWVD IHS变换 

分 类 号：TP391.41[自动化与计算机技术—计算机应用技术]