Surface Sensing & Monitoring SessionsNo Access

SUPPORT VECTOR CLASSIFICATION OF LAND COVER AND BENTHIC HABITAT FROM HYPERSPECTRAL IMAGES

VIDYA MANIAN

Department of Electrical and Computer Engineering, University of Puerto Rico, Mayaguez, PR 00682-6242, USA

Search for more papers by this author

and

MIGUEL VELEZ-REYES

Department of Electrical and Computer Engineering, University of Puerto Rico, Mayaguez, PR 00682-6242, USA

Search for more papers by this author

https://doi.org/10.1142/S0129156408005382Cited by:3 (Source: Crossref)

Abstract

This paper presents a novel wavelet and support vector machine (SVM) based method for hyperspectral image classification. A 1-D wavelet transform is applied to the pixel spectra, followed by feature extraction and SVM classification. Contrary to the traditional method of using pixel spectra with SVM classifier, our approach not only reduces the dimension of the input pixel feature vector but also improves the classification accuracy. Texture energy features computed in the spectral dimension are mapped using polynomial kernels and used for training the SVM classifier. Results with AVIRIS and other hyperspectral images for land cover and benthic habitat classification are presented. The accuracy of the method with limited training sets and computational burden is assessed.

Keywords:

Remember to check out the Most Cited Articles!
Check out these Notable Titles in Antennas

References

D. A. Landgrebe , Signal theory methods in multispectral remote sensing ( Wiley , NJ , 2003 ) . Crossref, Google Scholar
J. A. Gualtieri, Hyperspectral analysis, the support vector machine, and land and benthic habitats, Proceedings of IEEE workshop on Advances in techniques for analysis of remotely sensed data (2003) pp. 354–363. Google Scholar
M. Lennon, G. Mercier and L. Hubert-Moy, Classification of hyperspectral images with nonlinear filtering and SVMs, Proceedings IEEE Conf. IGARSS'023 pp. 1670–1672. Google Scholar
S. B. Serpico, M. D'Inca and G. Moser, Proceedings IGARSS'042 (2004) pp. 956–959. Google Scholar
M. F. Redondo, C. H. Espinosa and J. T. Sosperda, Proceedings IJCNN2 (2004) pp. 1145–1149. Google Scholar
G. C-Valls and L. Bruzzone, IEEE Trans GRS 43, 1351 (2005). Web of Science, Google Scholar
S. Kaewpijit, J. Le Moigne and T. El-Ghazawi, IEEE Trans. GRS (2003). Google Scholar
C. Liu , C. Zhao and W. Chen , Hyperspectral image classification by second generation wavelet based on adaptive band selection , Proceedings. IEEE Intl. Conf. Mechatronics and Automation ( 2005 ) . Google Scholar
S. Burrus , R. A. Gopinath and H. Gao , Introduction to wavelets and wavelet transforms-A primer ( Prentice Hall , 1998 ) . Google Scholar
I. Daubechies, IEEE Trans. Information Theory 32, 961 (1990). Web of Science, Google Scholar
C. J. Burges, Data mining and knowledge discovery, ed. U. Fayyad (Kluwer Academic, 1998) pp. 1–42. Google Scholar
V. N. Vapnick , Statistical learning theory ( John wiley and sons , 1998 ) . Google Scholar
OSU SVM Classifier Matlab Toolbox (ver 3.00) http://www.ece.osu.edu/~maj/osu_svm . Google Scholar
M. Sabri and J. Alirezai, Optimized space frequency kernel for texture classification, Proc. IEEE Intl. Conf. Image Processing3 (2004) pp. 1421–1524. Google Scholar
A. Plazaet al., IEEE Trans. GARS 43, 466 (2004). Google Scholar
C. Kotropoulos and I. Pitas , IEEE Trans. Signal Processing 42 , 2581 . Crossref, Web of Science, Google Scholar