inria-00492702, version 1
Classification de formulaires manuscrits en-ligne à l'aide de réseaux bayésiens
Emilie Philippot
1Yolande Belaïd 1Abdel Belaïd a, 1
Colloque International Francophone sur l'Ecrit et le Document - CIFED 2010 (2010) 95-110
Résumé : In this paper, we propose a method of recognizing handwritten forms online using an electronic pen-type clip. This mode of writing leads unconstrained filling uncertain form and the loss of the type of the completed form. We propose a method to identify the class of the form based on the positioning of fields filled. Our approach relies on the use of Bayesian networks to exploit, using conditional probabilities, occurrences of filling some fields depending on others. The tests were conducted on the basis of 3200 forms provided by the company Actimage. The first experiments show an average recognition of 90%.
- a – UNIVERSITE NANCY 2
- 1 : READ (LORIA)
- INRIA – CNRS : UMR7503 – Université Henri Poincaré - Nancy I – Université Nancy II – Institut National Polytechnique de Lorraine (INPL)
- Domaine : Informatique/Traitement du texte et du document
Informatique/Apprentissage - Mots-clés : Formulaires – écriture manuscrite en-ligne – classification – réseaux bayésiens
- inria-00492702, version 1
- http://hal.inria.fr/inria-00492702
- oai:hal.inria.fr:inria-00492702
- Contributeur : Yolande Belaid
- Soumis le : Mercredi 16 Juin 2010, 16:46:52
- Dernière modification le : Vendredi 15 Octobre 2010, 15:33:20
Documents associés
Exporter