For multitemporal hyperspectral images, the spectral characteristics of the same land cover object may vary significantly. Therefore, manifold alignment algorithm was employed to find a feature space in which data distributions of both images become the same. The method includes three steps. Firstly, a standard linear or nonlinear dimension reduction method is used to reduce the dimensionality of hyperspectral images. Secondly, the Procrustes analysis method is utilized to remove the translational, rotational and scaling components from one set so that the optimal alignment between the two data sets can be achieved. Finally, the nearest neighbor algorithm is applied for classification. Experimental results using multitemporal hyperion images demonstrate that the proposed approach can obtain performances which are superior to those of several popular manifold alignment methods.
鲁锦涛, 马丽. 基于流形对齐的高光谱遥感图像降维和分类算法[J]. 国土资源遥感, 2017, 29(1): 104-109.
LU Jintao, MA Li. Manifold alignment for dimension reduction and classification of multitemporal hyperspectral image. REMOTE SENSING FOR LAND & RESOURCES, 2017, 29(1): 104-109.
[1] Wang C.A Geometric Framework For Transfer Learning Using Manifold Alignment[D].Amherst,Massachusetts:University of Massachusetts Amherst,2010.
[2] Pan S J,Kwok J T,Yang Q.Transfer learning via dimensionality reduction[C]//Fox D,Gomes C P.Proceedings of the 23rd Conference on Artificial Intelligence.Chicago:AAAI Press,2008:677-682.
[3] Gu Q Q,Zhou J.Learning the shared subspace for multi-task clustering and transductive transfer classification[C]//Proceedings of the Ninth IEEE international conference on data mining.Miami,FL:IEEE,2009:159-168.
[4] Blitzer J,McDonald R,Pereira F.Domain adaptation with structural correspondence learning[C]//Proceedings of the 2006 Conference on Empirical Methods in Natural Language Processing.Stroudsburg,PA,USA:Association for Computational Linguistics,2006:120-128.
[5] Kan M N,Wu J T,Shan S G,et al.Domain adaptation for face recognition:Targetize source domain bridged by common subspace[J].International Journal of Computer Vision,2014,109(1/2):94-109.
[6] Ham J,Lee D D,Saul L K.Semisupervised alignment of manifolds[C]//Ghahramani Z,Cowell R.Proceedings of the Annual Conference on Uncertainty in Artificial Intelligence. Edinburgh,Scotland,UK:UAI,2005,10:120-127.
[7] Lafon S,Keller Y,Coifman R R.Data fusion and multicue data matching by diffusion maps[J].IEEE Transactions on Pattern Analysis and Machine Intelligence,2006,28(11):1784-1797.
[8] Wang C,Mahadevan S.Manifold alignment using Procrustes analysis[C]//Proceedings of the 25th International Conference on Machine Learning.New York,NY,USA:ACM,2008:1120-1127.
[9] Yang H L,Crawford M M.Manifold alignment for multitemporal hyperspectral image classification[C]//Proceedings of 2011 IEEE International Geoscience and Remote Sensing Symposium(IGARSS).Vancouver,BC,Canada:IEEE,2011:4332-4335.
[10] Yang H L,Crawford M M.Learning a joint manifold with global-local preservation for multitemporal hyperspectral image classification[C]//Proceedings of 2013 IEEE International Geoscience and Remote Sensing Symposium(IGARSS).Melbourne,Australia:IEEE,2013:1047-1050.
[11] Tuia D,Volpi M,Trolliet M,et al.Semisupervised manifold alignment of multimodal remote sensing images[J].IEEE Transactions on Geoscience and Remote Sensing,2014,52(12):7708-7720.
[12] Yan S C,Xu D,Zhang B Y,et al.Graph embedding and extensions:A general framework for dimensionality reduction[J].IEEE Transactions on Pattern Analysis and Machine Intelligence,2007,29(1):40-51.
[13] Yan S C,Xu D,Zhang B Y,et al.Graph embedding:A general framework for dimensionality reduction[C]//Proceedings of IEEE Computer Society Conference on Computer Vision and Pattern Recognition.San Diego,CA,USA:IEEE,2005,2:830-837.
[14] Sibson R.Studies in the robustness of multidimensional scaling:Procrustes statistics[J].Journal of the Royal Statistical Society.Series B(Methodological),1978,40(2):234-238.
2017, Vol. 29 
