Abstract
Handwriting recognition (HR) has always been a challenging problem for the Artificial Intelligence community, and remains an open issue. Given the complexity of the HR task (different writing and character styles, and writing conditions) it is perhaps not surprising that most of the success has stemmed from the development of task-specific HR systems. In this paper, we describe the Scribble system for automatically configuring HR systems (from a library of basic components) for well defined form-recognition tasks. The Scribble system is novel in that it integrates form design software with an automatic HR configuration system. The form designer not only allows a user to design a form, but also provides a means of capturing semantic information about the form’s fields (type, location, input constraints etc.), which it uses to guide the configuration process. The result is a specialised HR system that has been customised for a particular form.
This is a preview of subscription content, log in via an institution to check access.
Access this chapter
Tax calculation will be finalised at checkout
Purchases are for personal use only
Preview
Unable to display preview. Download preview PDF.
References
Dzuba, G., Filatov, A., Gershuny, D., Kil, I, Nikitin, V.: Check amount recognition based on the cross validation of courtesy and legal amount fields. Int. Journal of Pattern Recognition and A.I., vol 11 no. 4, pp 639–655, 1997.
Lam, L., Suen, C.Y., Guillevic, D., Strathy, N.W., Cheriet, M., Liu, K., Said J.N.: Automatic Processing of Information on Cheques. Int., Conf., on Systems, Man & Cybernetics, Vancouver, Canada, pp. 2353–2358, 1995.
Mahadevan, U., Srihari, S. N.: Hypothesis Generation for Word Separation in Handwritten Lines. Progress in Handwriting Recognition, proceedings of the IWFHR5, pp. 515–518, 1996.
O’Boyle, C., Smyth, B., and Geiselbrechtinger, F.: Scribble: Configuring Hand-Writing Recognition Systems from Form Knowledge. Proceedings of the 10th Irish Conference on Artificial Intelligence and Cognitive Science, pp. 217–222, Cork, Ireland, September 1999.
Parui, S. K., Chaudhuri, B. B., Majumder, D. D.: A Procedure for Recognition of Connected Handwritten Numerals. Int. J. Systems Sci., vol. 13, no. 9, pp 1019–1029, 1982.
Simoncini, L., Kovács-V, Zs. M.: A System for reading USA Census’ 90 Handwritten Fields. D.E.I.S., Int. Conf. on Document Analysis and Recognition, pp. 82–85, 1995.
Srihari, S. N, Shin, Y. C., Ramanaprasad, V., Lee, D. S.: Name and Address Block Reader. Proceedings of IEEE, vol. 84(7), 1996, pp. 1038–1049.
Wang, P.S.P, Nagendraprasad, M.V., Gupta, A.: A neural net based hybrid approach to handwritten numeral recognition. From Pixels to features III: Frontiers in Handwriting Recognition, (IWFHR’ 92), Elsevier Science Publishers, 1992.
Yanikoglu, Berrin A., Sandon, Peter A.: Off-line Cursive Handwriting Recognition Using Style Parameters. Technical Report, Department of Mathematics and Computer Science, Dartmouth College, Hanover, NH, 03755, June 7, 1993.
Author information
Authors and Affiliations
Editor information
Editors and Affiliations
Rights and permissions
Copyright information
© 2000 Springer-Verlag Berlin Heidelberg
About this paper
Cite this paper
O’Boyle, C., Smyth, B., Geiselbrechtinger, F. (2000). An Automatic Configuration System for Handwriting Recognition Problems. In: Logananthara, R., Palm, G., Ali, M. (eds) Intelligent Problem Solving. Methodologies and Approaches. IEA/AIE 2000. Lecture Notes in Computer Science(), vol 1821. Springer, Berlin, Heidelberg. https://doi.org/10.1007/3-540-45049-1_61
Download citation
DOI: https://doi.org/10.1007/3-540-45049-1_61
Published:
Publisher Name: Springer, Berlin, Heidelberg
Print ISBN: 978-3-540-67689-8
Online ISBN: 978-3-540-45049-8
eBook Packages: Springer Book Archive