An Integrated System for the Analysis and the Recognition of Characters in Ancient Documents

  • Stefano Vezzosi
  • Luigi Bedini
  • Anna Tonazzini
Conference paper
Part of the Lecture Notes in Computer Science book series (LNCS, volume 2423)


This paper describes an integrated system for processing and analyzing highly degraded ancient printed documents. For each page, the system reduces noise by wavelet-based filtering, extracts and segments the text lines into characters by a fast adaptive thresholding, and performs OCR by a feed-forward back-propagation multilayer neural network. The probability recognition is used as a discriminant parameter for determining the automatic activation of a feed-back process, leading back to a block for refining segmentation. This block acts only on the small portions of the text where the recognition was not trustable, and makes use of blind deconvolution and MRF-based segmentation techniques. The experimental results highlight the good performance of the whole system in the analysis of even strongly degraded texts.


Markov Random Fields Text Line Optical Character Recognition Gray Level Image Character Segmentation 
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.


  1. 1.
    Donoho, D.L.: IEEE Trans. Information Theory. 41 (1995) 613–627.zbMATHCrossRefMathSciNetGoogle Scholar
  2. 2.
    Vogl, T.P. et al.: Biological Cybernetics. 59 (1988) 256–264.CrossRefGoogle Scholar
  3. 3.
    Kundur, D., Hatzinakos, D.: IEEE Sig. Proc. Mag. (1996) 43–62.Google Scholar
  4. 4.
    Li, S.Z.: Markov Random Field Modeling in Computer Vision. (1995) Springer-Verlag Tokyo.Google Scholar
  5. 5.
    Tonazzini, A., Bedini, L.: Proc. 10th ICIAP. (1999) 836–841.Google Scholar
  6. 6.
    Ayers, G.R., Dainty, J.G.: Opt. Lett. 13 (1988) 547–549.CrossRefGoogle Scholar
  7. 7.
    Aarts, E., Korst, J.: Simulated Annealing and Boltzmann Machines. (1989) Wiley.Google Scholar

Copyright information

© Springer-Verlag Berlin Heidelberg 2002

Authors and Affiliations

  • Stefano Vezzosi
    • 1
  • Luigi Bedini
    • 1
  • Anna Tonazzini
    • 1
  1. 1.CNRIstituto di Elaborazione della InformazionePISAItaly

Personalised recommendations