A Gibbsian Kohonen Network for Online Arabic Character Recognition

  • Neila Mezghani
  • Amar Mitiche
Part of the Lecture Notes in Computer Science book series (LNCS, volume 5359)


The purpose of this study is to investigate handwritten online character recognition by Kohonen neural networks which learn class conditional Gibbs densities from training samples. The characters are represented by histograms (empirical distributions) of features. The Kohonen network learning algorithm implements a gradient ascent which maximizes an entropy criterion under constraints. Using a database of handwritten online Arabic characters produced without constraints by a large number of writers, we conducted extensive experiments which show the advantage of this Gibbsian Kohonen network over other classifiers such as a regular Kohonen neural network and a Gibbsian Bayes classifier.


Recognition Rate Character Recognition Feature Histogram Gradient Ascent Entropy Criterion 
These keywords were added by machine and not by the authors. This process is experimental and the keywords may be updated as the learning algorithm improves.


Unable to display preview. Download preview PDF.

Unable to display preview. Download preview PDF.


  1. 1.
    Al-Sheikh, T., El-Taweel, S.: Real-time Arabic handwritten character recognition. Pattern recognition 23, 1323–1332 (1990)CrossRefGoogle Scholar
  2. 2.
    Schenkel, M., Guyon, I., Henderson, D.: On-line script recognition using time delay neural networks and hidden markov models. In: International Conference on Acoustics, Speech, and Signal Processing, Adelaide, Austrailia, vol. 2, pp. 637–640 (1994)Google Scholar
  3. 3.
    Mezghani, N., Cheriet, M., Mitiche, A.: Combination of pruned kohonen maps for on-line Arabic characters recognition. In: Seventh International Conference on Document Analysis and Recognition, Edinburgh, Scotland, vol. 2, pp. 900–905 (2003)Google Scholar
  4. 4.
    Plamondon, R.: A model-based segmentation framework for computer processing of handwriting. In: Proceedings of the Eleventh IAPR International Conference on Pattern Recognition, pp. 303–307 (1992)Google Scholar
  5. 5.
    Kavianifar, M., Amin, A.: Preprocessing and structural feature extraction for a multi-fonts arabic/persian ocr. In: Proceedings of the Fifth International Conference on Document Analysis and Recognition, pp. 213–216 (1999)Google Scholar
  6. 6.
    Klassen, T.: Towards neural network recognition of handwritten Arabic letters. Master’s thesis, Dalhousie University, Halifax (2001)Google Scholar
  7. 7.
    Alimi, A.: A neuro-fuzzy approch to recognize on-line Arabic handwritting. In: International Conference on Neural Networks, vol. 3, pp. 1397–1400 (1997)Google Scholar
  8. 8.
    Mahmoud, S.A.: Arabic character-recognition using fourier descriptors and character contour encoding. Pattern Recognition 27, 815–824 (1994)CrossRefGoogle Scholar
  9. 9.
    Mezghani, N., Mitiche, A., Cheriet, M.: A new representation of character shape and its use in on-line character recognition by a self-organizing map. In: IEEE International Conference on Image Processing, ICIP, Singapore, vol. 3, pp. 2123–2126 (2004)Google Scholar
  10. 10.
    Zhu, S.C., Wu, Y., Mumford, D.: Minimax entropy principle and its application to texture modeling. Neural Comput. 9, 1627–1660 (1997)CrossRefGoogle Scholar
  11. 11.
    Zhu, S.: Embedding gestalt laws in markov random fields. IEEE Transactions on Pattern Analysis and Machine Intelligence 21, 1170–1187 (1999)CrossRefGoogle Scholar
  12. 12.
    Duda, P.O., Hart, P.E., Stork, D.G.: Pattern Classification, New York (2001)Google Scholar
  13. 13.
    Bail, E.L., Mitiche, A.: Quantification vectorielle d’images par le réseau neuronal de kohonen. Traitement du Signal 6, 529–539 (1989)Google Scholar
  14. 14.
    Anouar, F., Badran, F., Thiria, S.: Probabilistic self-organizing map and radial basis function. Journal of Neurocomputing 20, 83–96 (1998)CrossRefzbMATHGoogle Scholar
  15. 15.
    Mezghani, N., Mitiche, A., Cheriet, M.: Bayes classification of online arabic characters by gibbs modeling of class conditional densities. IEEE Transactions on Pattern Analysis and Machine Intelligence 30, 1121–1131 (2008)CrossRefGoogle Scholar

Copyright information

© Springer-Verlag Berlin Heidelberg 2008

Authors and Affiliations

  • Neila Mezghani
    • 1
  • Amar Mitiche
    • 2
  1. 1.École de technologie supérieureMontrealCanada
  2. 2.Institut national de la recherche scientifique, INRS-EMTMontrealCanada

Personalised recommendations