Combination of Linear Support Vector Machines and Linear Spectral Mixed Model for Spectral Unmixing
Abstract
Aiming at the shortcoming of linear spectral mixing model (LSMM) and linear support vector machines (LSVM) has potential capability to be used in spectral unmixing, but it is hard to construct too many classifiers when LSVM is used in partial unmixing. In this paper, the equality of LSVM and LSMM is proved concisely, and then a new double-unmixing scheme is proposed by combining the two models. In the first time, LSVM based full unmixing is performed for selecting related class subset. In the second time, appropriate model is selected according to the cardinality of current subset for partial unmixing. Another, least square LSVM has been improved for effective unmixing. Experiments prove the high efficiency of the proposed scheme.
Keywords
Fractional Image Linear Support Vector Machine Mixed Pixel Fractional Abundance Spectral Mixture AnalysisPreview
Unable to display preview. Download preview PDF.
References
- 1.Keshava, N., Mustard, J. F.: Spectral Unmixing. Signal Processing Magazine, IEEE. 19 (2002) 44–57CrossRefGoogle Scholar
- 2.Nielsen, A. A.: Spectral Mixture Analysis: Linear and Semi-Parametric Full and Iterated Partial Unmixing in Multi-and Hyperspectral Image Data. International Journal of Computer Vision, 42 (2001) 17–37MATHCrossRefGoogle Scholar
- 3.Daniel, N.: Evaluation of Stepwise Spectral Unmixing with Hydice Data. SIMG-503 Senior Research Final Report Center for Imaging Science. http://www.cis.rit.edu/research/thesis/bs/1999/newland/title.htmlGoogle Scholar
- 4.Martin, B., Hugh, G. L., Steve, R. G.: Linear Spectral Mixture Models and Support Vector Machines for Remote Sensing. IEEE Trans. on Geoscience and RemoteSensing, 38 (2000) 2346–2360CrossRefGoogle Scholar
- 5.Martin, B., Gunn, S. R., Lewis, H. G.: Support Vector Machines for Optimal Classification and Spectral Unmixing. Ecol. Modeling, 120 (1999) 167–179CrossRefGoogle Scholar
- 6.Vapnik, V. N.: The Nature of Statistical Learning Theory. New York, Springer Press, NY (1995)MATHGoogle Scholar
- 7.Wang, L.G., Zhang Y., Zhang, J.P.: A New Weighted Least Squares Support Vector Machines and Its Sequential Minimal Optimization Algorithm. Chinese J. Electron, 15 (2006), to be appear in No. 3Google Scholar
- 8.Hassan, E.: Introducing Correctness Coefficient as An Accuracy Measure for Sub Pixel Classification Results. www.ncc.org.ir/articles/poster83/H. Emami.pdfGoogle Scholar