Hyperspectral Data Selection from Mutual Information Between Image Bands

  • José Martínez Sotoca
  • Filiberto Pla
Part of the Lecture Notes in Computer Science book series (LNCS, volume 4109)


This work presents a band selection method for multi and hyperspectral images using correlation among bands based on mutual information measures. The relationship among bands are represented by means of the transinformation matrix. A process based on a Deterministic Annealing optimization is applied to the transinformation matrix in order to obtain a reduction of this matrix looking for the image bands as less uncorrelated as possible between them. Some experiments are presented to show the effectiveness of the bands selected from the point of view of pixel classification.


Multispectral images mutual information deterministic annealing unsupervised feature selection 


  1. 1.
    Aczel, J., Daroczy, Z.: On measures of information and their characterization. Academic Press, New York (1975)Google Scholar
  2. 2.
    Agrawal, R., Gehrke, J., Gunopulos, D., Raghavan, P.: Automatic subspace clustering of high dimensional for data mining applications. In: Proceedings of the ACM SIGMOD International Conference on Management of Data, Seattle, WA, pp. 94–105 (June 1998)Google Scholar
  3. 3.
    Bruzzonne, L., Roli, F., Serpico, S.B.: An extension to multiclass cases of the Jeffreys-Matusita distance. IEEE Transactions on Geoscience and Remote Sensing 33, 1318–1321 (1995)CrossRefGoogle Scholar
  4. 4.
    Groves, P., Bajcsy, P.: Methodology for hyperspectral band and classification model selection. In: Landgrebe, D.A. (ed.) IEEE Workshop on Advances in Techniques for Analysis of Remotely Sensed Data, Washington D.C. (2003)Google Scholar
  5. 5.
    Jaynes, E.T.: Prior Probatilities. IEEE Transactions on System Science and Cybernetic SSC-4, 227–241 (1968); Concepts and Applications of Modern Decision Models. In: Tummala, V.M.R., Henshaw, R.C. (eds.) Michigan State University Business Studies Series (reprinted 1976) Google Scholar
  6. 6.
    Kononenko, I.: Estimating attributes: analysis and extensions of RELIEF. In: Proceedings of 7th European Conference on Machine Learning, Catania, Italy, pp. 171–182 (1994)Google Scholar
  7. 7.
    Kumar, S., Ghosh, J., Crawford, M.M.: Best basis feature extraction algorithms for classification of hyperspectral data. IEEE Transactions on Geoscience and Remote Sensing 39(7), 1368–1379 (2001)CrossRefGoogle Scholar
  8. 8.
    Sotoca, J.M., Pla, F., Klaren, A.C.: Unsupervised band selection for multispectral images using information theory. In: 17th International Conference on Pattern Recognition, Cambridge (UK), vol. 3, pp. 510–513 (2004)Google Scholar

Copyright information

© Springer-Verlag Berlin Heidelberg 2006

Authors and Affiliations

  • José Martínez Sotoca
    • 1
  • Filiberto Pla
    • 1
  1. 1.Dept. Llenguatges i Sistemes InformáticsUniversitat Jaume ICastellóSpain

Personalised recommendations