A l1-minimization Based Approach for Hyperspectral Data Classification



Most of classification methods require model learning procedure or optimal parameters selection using large number of training samples. In this paper, we propose a novel classification approach using l1-minimization based sparse representation which does not need any learning procedure or parameters selection. The proposed approach is based on l1minimization because l0-minimization is generally NP-hard and is not a convex optimization problem. Sparse based solutions have been proposed in other areas like signal processing but to the best of our knowledge, this is the first time that sparse based classification is being proposed for the classification of hyperspectral data. l1-minimization based sparse representation is calculated for each of the test samples using few training samples directly. In remote sensing, usually it is extremely difficult and expensive to identify and label the samples due to which we often lack sufficient training samples for classification. We tested the proposed approach for difficult classification problem i.e high dimensional spaces and few training samples. We also analyzed the cases in which the sparse based classification may and may not work. Our extensive experiments on real hyperspectral dataset (AVIRIS 1992 Indiana’s Indian Pines image), prove that the proposed technique offers more classification accuracy and more efficient than state-of-the-art SVM and semi-supervised methods as it does not need any model selection.




David Wang




S. U. H. Qazi et al., "A l1-minimization Based Approach for Hyperspectral Data Classification", Key Engineering Materials, Vol. 500, pp. 675-681, 2012


January 2012




[1] G. Camps-Valls, T. Bandos, and D. Zhou. Semi-supervised graph-based hyperspectral image classification. IEEE Transactions on Geoscience and Remote Sensing, 45(10): 2044–3054, (2007).

DOI: https://doi.org/10.1109/tgrs.2007.895416

[2] E. Candes and J. Romberg. l1magic: a collection of matlab routines for solving the convex optimization programs central to compressive sampling, (2006).

[3] E. J. Candes, J. Romberg, and T. Tao. Robust uncertainty principles: exact signal reconstruction from highly incomplete frequency information. IEEE Transactions on Information Theory, 52(2): 489–509, (2006).

DOI: https://doi.org/10.1109/tit.2005.862083

[4] E. J. Candes and T. Tao. Near-optimal signal recovery from random projections: universal encoding strategies? IEEE Transactions on Information Theory, 52(12): 5406–5425, (2006).

DOI: https://doi.org/10.1109/tit.2006.885507

[5] D. L. Civco. Artificial neural networks for landcover classification and mapping. International Journal of Geophysical Information Systems, 7(2): 173–186, (1993).

[6] D. L. Donoho. Compressed sensing. IEEE Transactions on Information Theory, 52(4): 1289–1306, (2006).

DOI: https://doi.org/10.1109/tit.2006.871582

[7] J. A. Gualtieri, S. R. Chettri, R. F. Cromp, and L. F. Johnson. Support vector machine classifiers as applied to aviris data. In in Proceedings of the 8th JPL Airborne Geoscience Workshop, Feb. (1999).

[8] M. A. Leel, S. Prasad, L. M. Bruce, T. R. West, D. Reynolds, T. Irby, and H. Kalluril. Sensitivity of hyperspectral classification algorithms to training sample size. IEEE Geoscience and Remote Sensing Letters, 55(1): 119–139, (2009).

DOI: https://doi.org/10.1109/whispers.2009.5288983

[9] B. Scholkopf and A. Smola. Learning With Kernels Support Vector Machines, Regularization, Optimization and Beyond. MIT Press Series, Cambridge, MA, (2002).

DOI: https://doi.org/10.1145/1027914.1027921

[10] R. Tibshirani. Regression shrinkage and selection via the lasso. Journal of the Royal Statistical Society. Series B, 58(1): 267–288, (1996).

[11] J. Wright, A. Yang, A. Ganesh, S. Sastry, and Y. Ma. Robust face recognition via sparse representation. IEEE Trans. on Pattern Analysis and Machine Intelligence, 31(2): 210–227, (2009).

DOI: https://doi.org/10.1109/tpami.2008.79

[12] H. Wu, Kuang, Gangyao, and W. Yu. Unsupervised classification method for hyperspectral image combining pca and gaussian mixture model. Journal of Computer and System Sciences, 5286: 729–734, (2003).

DOI: https://doi.org/10.1117/12.539872