一种改进的大尺度高光谱流形降维算法

张晶晶; 周晓勇; 刘奇

doi:doi:10.3788/aos201333.1128001

光学学报, 2013, 33 (11): 1128001, 网络出版: 2013-10-20

一种改进的大尺度高光谱流形降维算法

Improved Dimensionality Reduction Algorithm of Large-Scale Hyperspectral Scenes Using Manifold

论文大纲

张晶晶 ^1,2,*周晓勇 ¹刘奇 ¹

作者单位

¹ 安徽大学电气工程与自动化学院, 安徽合肥 230601

² 中国科学院通用光学定标与表征技术重点实验室, 安徽合肥 230031

摘要

经典流形算法等距映射（ISOMAP）和局部线性嵌入（LLE）可以对高光谱数据进行降维，但不能解决大尺度高光谱图像的流形降维难题。详细论述了ISOMAP和LLE在大尺度高光谱流形降维中遇到的问题，提出了一种基于增量等距映射（IISOMAP）和LLE结合的高光谱流形降维算法IISOMAP-LLE，并针对流形降维算法较线性降维算法最小噪声分离（MNF）可以更好地发掘出高光谱数据中的非线性结构的优点，通过AVIRIS和OMIS-II数据实验验证了算法的可行性和优越性，并证明了IISOMAP-LLE算法可以避免增强型等距映射(ENH-ISOMAP)中由于Landmark点选取不当而造成的本征维数较大时类间可分性反而低于MNF的缺点。

Abstract

It is practicable for dimensionality reduction of hyperspectral scenes using manifold algorithm such as isometric mapping (ISOMAP) and local linear embedding (LLE). However the two classical manifold algorithm are not suitable for solving the large-scale hyperspectral scenes. We elaborate the problems encountered in applying ISOMAP and LLE to dimensionality reduction of large-scale hyperspectral scenes, then an improved algorithm called IISOMAP-LLE, which is based on incremental isometric mapping (IISOMAP) and LLE, is proposed to represent the nonlinear structure of hyperspectral imagery that linear algorithm minimum noise fraction (MNF) could not discover. At last we demonstrate two experiments using large-scale AVIRIS and OMIS-II hyperspectral scenes to illustrate the approach. Experimental results prove that the IISOMAP-LLE not only is much better than linear algorithm MNF but also can avoid superiority decline of separability compared with MNF that encounterd in enhanced isometric mapping.

参考文献

[1] 童庆禧, 张兵, 郑兰芬. 高光谱遥感[M]. 北京: 高等教育出版社, 2006.

Tong Qingxi, Zhang Bing, Zheng Lanfen. Hyperspectral Remote Sensing [M]. Beijing: China Higher Education Press, 2006.

[2] A N Gorban, Balazs Kegl, D C WunschII, et al.. Principal Manifolds for Data Visualization and Dimension Reduction [M]. Berlin: Springer-Verlag Berlin and Heideberg GmbH & Co. K, 2007.

[3] 杜培军, 王小美, 谭琨, 等. 利用流形学习进行高光谱遥感影像的降维与特征提取[J]. 武汉大学学报·信息科学版, 2011, 36(2): 148-152.

Du Peijun, Wang Xiaomei, Tan Kun, et al.. Dimensionality reduction and feature extraction from hyperspectral remote sensing imagery based on manifold learning [J]. Geomatics and Information Science of Wuhan University, 2011, 36(2): 148-152.

[4] J B Tenenbaum, V De Silva, J C Langford. A global geometric framework for nonlinear dimensionality reduction [J]. Science, 2000, 290(5500): 2319-2323.

[5] S T Roweis, L K Saul. Nonlinear dimensionality reduction by locally linear embedding [J]. Science, 2000, 290(5500): 2323-2326.

[6] Y Chen, M M Crawford, J Ghosh. Applying nonlinear manifold learning to hyperspectral data for land cover classification [C]. International Geoscience and Remote Sensing Symposium, Seoul, 2005, 6: 4311.

[7] D Guangjun, Z Yongsheng, J Song. Dimensionality reduction of hyperspectral data based on ISOMAP algorithm [C]. ICEMI′07.8th International Conference on Electronic Measurement and Instruments, IEEE, Xi′an，2007: 3-935-3-938.

[8] X R Wang, S Kumar, T Kaupp, et al.. Applying ISOMAP to the learning of hyperspectral image [C]. Proceedings of the 2005 Australasian Conference on Robotics & Automation. Australian Robotics & Automation Association, Sydney, 2005.

[9] C M Bachmann, T L Ainsworth, R A Fusina, et al.. Manifold coordinate representations of hyperspectral imagery: improvements in algorithm performance and computational efficiency [C]. Geoscience and Remote Sensing Symposium (IGARSS), 2010 IEEE International. IEEE, Honolulu, 2010: 4244-4247.

[10] C M Bachmann, T L Ainsworth, R A Fusina. Exploiting manifold geometry in hyperspectral imagery [J]. IEEE Transactions on Geoscience and Remote Sensing, 2005, 43(3): 441-454.

[11] C M Bachmann, T L Ainsworth, R A Fusina. Improved manifold coordinate representations of large-scale hyperspectral scenes [J]. IEEE Transactions on Geoscience and Remote Sensing, 2006, 44(10): 2786-2803.

[12] M H C Law, A K Jain. Incremental nonlinear dimensionality reduction by manifold learning [J]. Pattern Analysis and Machine Intelligence, IEEE Transactions on, 2006, 28(3): 377-391.

[13] H C Law. Clustering, Dimensionality Reduction, and Side Information [D]. Michigan: Michigan State University, 2006.

[14] D Zhao, L Yang. Incremental isometric embedding of high-dimensional data using connected neighborhood graphs [J]. Pattern Analysis and Machine Intelligence, IEEE Transactions on, 2009, 31(1): 86-98.

[15] D K Agrafiotis, H Xu. A self-organizing principle for learning nonlinear manifolds [J]. Proceedings of the National Academy of Sciences, 2002, 99(25): 15869-15872.

[16] V A Tolpekin, A Stein. Quantification of the effects of land-cover-class spectral separability on the accuracy of Markov-random-field-based superresolution mapping [J]. Geoscience and Remote Sensing, IEEE Transactions on, 2009, 47(9): 3283-3297.

[17] T Murakami, S Ogawa, N Ishitsuka, et al.. Crop discrimination with multitemporal SPOT/HRV data in the Saga Plains, Japan [J]. International Journal of Remote Sensing, 2001, 22(7): 1335-1348.

[18] S M Davis, D A Landgrebe, T L Phillips, et al.. Remote sensing: the quantitative approach [J]. New York: McGraw-Hill International Book Co., 1978. 405 p., 1978, 1.

[19] M L Davison. Multidimensional Scaling [M]. New York: Wiley, 1983.

[20] V Silva, J B Tenenbaum. Global versus local methods in nonlinear dimensionality reduction [J]. Advances in Neural Information Processing Systems, 2002, 15: 705-712.

[21] L K Saul, S T Roweis. Think globally, fit locally: unsupervised learning of low dimensional manifolds [J]. The Journal of Machine Learning Research, 2003, 4: 119-155.

[22] T Friedrich. Nonlinear dimensionality reduction-locally linear embedding versus isomap [C]. The University of Sheffield, 2004.

[23] M Bernstein, V De Silva, J C Langford, et al.. Graph Approximations to Geodesics on Embedded Manifolds [R]. Stanford: Stanford University, 2000.

[24] E Levina, P J Bickel. Maximum likelihood estimation of intrinsic dimension [J]. Ann Arbor MI, 2004, 48109: 1092.

张晶晶, 周晓勇, 刘奇. 一种改进的大尺度高光谱流形降维算法[J]. 光学学报, 2013, 33(11): 1128001. Zhang Jingjing, Zhou Xiaoyong, Liu Qi. Improved Dimensionality Reduction Algorithm of Large-Scale Hyperspectral Scenes Using Manifold[J]. Acta Optica Sinica, 2013, 33(11): 1128001.

一种改进的大尺度高光谱流形降维算法

关于本站 Cookie 的使用提示

全站搜索

一种改进的大尺度高光谱流形降维算法

相关论文

相关资讯

关于本站 Cookie 的使用提示

全站搜索