Relevance-Based Feature Extraction for Hyperspectral Images

Hyperspectral imagery affords researchers all discriminating details needed for fine delineation of many material classes. This delineation is essential for scientific research ranging from geologic to environmental impact studies. In a data mining scenario, one cannot blindly discard information because it can destroy discovery potential. In a supervised classification scenario, however, the preselection of classes presents one with an opportunity to extract a reduced set of meaningful features without degrading classification performance. Given the complex correlations found in hyperspectral data and the potentially large number of classes, meaningful feature extraction is a difficult task. We turn to the recent neural paradigm of generalized relevance learning vector quantization (GRLVQ) [B. Hammer and T. Villmann, Neural Networks, vol. 15, pp. 1059-1068, 2002], which is based on, and substantially extends, learning vector quantization (LVQ) [T. Kohonen, Self-Organizing Maps, Berlin, Germany: Springer-Verlag, 2001] by learning relevant input dimensions while incorporating classification accuracy in the cost function. By addressing deficiencies in GRLVQ, we produce an improved version, GRLVQI, which is an effective analysis tool for high-dimensional data such as remotely sensed hyperspectral data. With an independent classifier, we show that the spectral features deemed relevant by our improved GRLVQI result in a better classification for a predefined set of surface materials than using all available spectral channels.

[1]  B. Ripley,et al.  Pattern Recognition , 1968, Nature.

[2]  Duane DeSieno,et al.  Adding a conscience to competitive learning , 1988, IEEE 1988 International Conference on Neural Networks.

[3]  Stanley C. Ahalt,et al.  Competitive learning algorithms for vector quantization , 1990, Neural Networks.

[4]  Franck Poirier,et al.  DVQ: DYNAMIC VECTOR QUANTIZATION - AN INCREMENTAL LVQ , 1991 .

[5]  Biing-Hwang Juang,et al.  Discriminative learning for minimum error classification [pattern recognition] , 1992, IEEE Trans. Signal Process..

[6]  Erzsébet Merényi,et al.  Classification of the LCVF AVIRIS test site with a Kohonen artificial neural network , 1993 .

[7]  Atsushi Sato,et al.  Generalized Learning Vector Quantization , 1995, NIPS.

[8]  R. Gray,et al.  Combining Image Compression and Classification Using Vector Quantization , 1995, IEEE Trans. Pattern Anal. Mach. Intell..

[9]  Erzsebet Merenyi,et al.  Classification of hyperspectral images using wavelet transforms and neural networks , 1995, Optics + Photonics.

[10]  Jeffrey S. Miller,et al.  Mapping of Spectral Variations on the Surface of Mars from High Spectral Resolution Telescopic Images , 1996 .

[11]  Atsushi Sato,et al.  A formulation of learning vector quantization using a new misclassification measure , 1998, Proceedings. Fourteenth International Conference on Pattern Recognition (Cat. No.98EX170).

[12]  T. Minor,et al.  QUANTITATIVE COMPARISON OF NEURAL NETWORK AND CONVENTIONAL CLASSIFIERS FOR HYPERSPECTRAL IMAGERY , 1998 .

[13]  Atsushi Sato.  An analysis of initial state dependence in generalized LVQ , 1999 .

[14]  Mu-Chun Su,et al.  An efficient initialization scheme for the self-organizing feature map algorithm , 1999, IJCNN'99. International Joint Conference on Neural Networks. Proceedings (Cat. No.99CH36339).

[15]  David G. Stork,et al.  Pattern Classification (2nd ed.) , 1999 .

[16]  Peter Sinc̆ák,et al.  Quo vadis computational intelligence? : new trends and approaches in computational intelligence , 2000 .

[17]  Peter L. Bartlett,et al.  Rademacher and Gaussian Complexities: Risk Bounds and Structural Results , 2003, J. Mach. Learn. Res..

[18]  Shigeo Abe DrEng.  Pattern Classification , 2001, Springer London.

[19]  Thomas Villmann,et al.  Learning Vector Quantization for Multimodal Data , 2002, ICANN.

[20]  Koby Crammer,et al.  Margin Analysis of the LVQ Algorithm , 2002, NIPS.

[21]  Thomas Villmann,et al.  Generalized relevance learning vector quantization , 2002, Neural Networks.

[22]  E. Mer,et al.  \Precision Mining" of High-Dimensional Patterns with Self-Organizing Maps: Interpretation of Hyperspectral Images Erzs , 2003 .

[23]  Thomas Villmann,et al.  Neural maps in remote sensing image analysis , 2003, Neural Networks.

[24]  Thomas Villmann,et al.  Supervised Neural Gas with General Similarity Measure , 2005, Neural Processing Letters.

[25]  Thomas Villmann,et al.  On the Generalization Ability of GRLVQ Networks , 2005, Neural Processing Letters.

[26]  Nicolas H. Younan,et al.  Wavelet domain statistical hyperspectral soil texture classification , 2005, IEEE Transactions on Geoscience and Remote Sensing.

[27]  M. V. Velzen,et al.  Self-organizing maps , 2007 .

[28]  S. Katagiri,et al.  Discriminative Learning for Minimum Error Classification , 2009 .