两种基于2DPCA图像融合算法的性能研究Performance of two kinds of 2DPCA algorithms in image fusion
吴学明,杨武年
摘要(Abstract)：
基于2DPCA的图像融合算法是一种新近提出的图像融合技术,它是通过图像矩阵右乘投影矩阵来实现2DPCA分析的。因而,这种算法能够有效利用图像的二维结构信息,具有良好的性能,不仅能够提高融合图像的空间分辨率,而且使光谱信息得到较好保持。提出了一种通过图像矩阵左乘最优投影矩阵来实现2DPCA分析的图像融合算法,即L2DPCA图像融合算法,并将2DPCA算法和L2DPCA算法与经典的BROVEY、PCA和H IS三种图像融合算法进行了对比研究。结果表明:2DPCA算法和L2DPCA算法的性能明显优于这三种经典的算法。
关键词(KeyWords)： 图像融合;遥感;2DPCA;L2DPCA
基金项目(Foundation):
作者(Author): 吴学明,杨武年
参考文献(References)：
- [1]Henri,Maiter,Isabelle Bloch.Image Fusion[J].Vis-tas in Astronomy,1997,41(3):329-335.
- [2]Towinn,Taxt,Anne H.Schistad Solberg.InformationFusion in Remote Sensing[J].Vistas in Astronomy,1997,41(3):337-342.
- [3]C.POHL,J.L.VAN GENDEREN.Multisensor imagein remote sensing:concepts,methods and applications[J].International Journal of Remote Sensing,1998,19(5):823-854.
- [4]YANG Jian,David ZHANG,et al.Two-Dimensional PCA:A NewApproach to Appearance-Based Face Representationand Recognition[J].IEEE Transactions on Pattern Analy-sis and Machine Intelligence,2004,26(1).
- [5]Daoqiang Zhang,Zhihua Zhou.(2D)2PCA:Two-direc-tional two-dimensional PCAfor efficient face repre-senta-tion and recognition.[J].Neurocomputing 2005,69.
- [6]KONG Hui,LI Xu-chun,et al.Generalized 2D PrincipalComponent Analysis[C]//International Joint Confer-ence on Neural Networks,Montrial,Canada,2005.
- [7]KONG Hui,WANG Lei,et al.Generalized 2D PrincipalComponent Analysis for Face Image Representation andRecognition[J].Neural Networks 2005,18:585?594.
- [8]D.S.Guru,P.Nagabhushan and B.H.Shekar.Two Di-mensional Principal Component Analysis for VisualLearning and Recognition of 3D Objects:A Robust andan Efficient Approach[EB/OL].Proceedings of the In-terna-tional Conference on Cognition and Recognition.http://www.studentprogress.com/appln/colleges/cogrec/Papers/C_06.pdf,[2007-04-06].
- [9]Wu Xue-ming,YANG Wu-nian.A novel image fusionmethod based on2DPCAin remote sensing[C]//Interna-tional Archives of Photogrammetry,Remote Sensing andSpatial Information Sciences.Vol.XXXVI,Part 7/C54,Comm.VII,ISPRS,Chengdu,China,2007:215-218.