Perceptual features for off-line handwritten word recognition: A framework for heuristic prediction, representation and matching

  • Sriganesh Madhvanath
  • Venu Govindaraju
Poster Papers
Part of the Lecture Notes in Computer Science book series (LNCS, volume 1451)


Perceptual holistic features are visually conspicuous features of the word shape that have been cited in reading studies as being utilized in fluent reading. While these features have been used for word recognition when the lexicon of possible words is small and static, their application to the general problem of omni-scriptor handwritten word recognition is limited by their variability at the word level and the paucity of samples for word-level training. A methodology of coarse holistic features and heuristic prediction of ideal features from ASCII is proposed to address these issues. The methodology is based on the view that real world examples of handwritten words are instances of the ideal exemplar of the word class distorted by the scriptor, stylus, medium and intervening electronic imaging processes, and has applications in verification and lexicon reduction for handwritten word recognition.


holistic approach handwritten word recognition perceptual features graph-matching handwriting models syntactic approach 


  1. 1.
    Soltysiak, S.J.: Visual information in word recognition: Word shape or letter identities? In: Proc. Wkshp on Integration of Nat. Lang. and Vision Proc., Seattle, USA, Aug. 2–3 (1994)Google Scholar
  2. 2.
    Lecolinet, E., Baret, O.: Cursive word recognition: Methods and strategies. In: Impedovo, S. (ed.): Fundamentals in Handwriting Recognition. Springer-Verlag (1993) 235–263Google Scholar
  3. 3.
    Paquet, T., Lecourtier, Y.: Handwriting recognition: Application to bank cheques. In: Proc. First Intl. Conf. Document Anal. Recog., Saint-Malo, France, Sept. (1991) 749–757Google Scholar
  4. 4.
    Moreau, J.V.: A new system for automatic reading of postal checks. In: Impedovo, S., Simon, J.C. (eds): From Pixels to Features III: Frontiers in Handwriting Recognition. North-Holland (1992) 171–184Google Scholar
  5. 5.
    Simon, J.C., Baret, O., Gorski, N.D.: A system for the recognition of handwritten literal amounts of checks. In: Spitz, A.L., Dengel, A. (eds): Proc. IAPR Wkshp Document Anal. Sys. World Scientific (1994) 265–287Google Scholar
  6. 6.
    Guillevic, D., Suen, C.Y.: Cursive script recognition applied to the processing of bank cheques. In: Proc. Third Intl. Conf. Document Anal. Recog., Montreal, Canada, Aug. 14–16. IEEE Computer Society (1995) 11–14Google Scholar
  7. 7.
    Olivier, C., Paquet, T., Avila, M., Lecourtier, Y.: Recognition of handwritten words using stochastic models. In: Proc. Third Intl. Conf. Document Anal. Recog., Montreal, Canada, Aug. 14–16. IEEE Computer Society (1995) 19–23Google Scholar
  8. 8.
    Madhvanath, S., Govindaraju, V.: Holistic lexicon reduction. In: Proc. Third Intl. Wkshp. Frontiers in Handwriting Recog., Buffalo, USA, May 25–27 (1993) 71–81Google Scholar
  9. 9.
    Madhvanath, S., Kleinberg, E., Govindaraju, V., Srihari, S.N.: The HOVER system for rapid holistic verification of off-line handwritten phrases. In: Proc. Fourth Intl. Conf. Document Anal. Recog., Ulm, Germany, Aug. 18–20. IEEE Computer Society (1997) 855–859Google Scholar

Copyright information

© Springer-Verlag Berlin Heidelberg 1998

Authors and Affiliations

  • Sriganesh Madhvanath
    • 1
  • Venu Govindaraju
    • 2
  1. 1.IBM Almaden Research CenterSan JoseUSA
  2. 2.Center of Excellence for Document Analysis and Recognition (CEDAR) Department of Computer ScienceState University of New York at BuffaloBuffaloUSA

Personalised recommendations