Handwritten text recognition through writer adaptation

A. Nosary1, T. Paquet1, L. Heutte1, A. Bensefia1
1Laboratoire Perception Systèmes InformationUFR des Sciences, Université de Rouen, Mont Saint Aignan, France

Tóm tắt

Handwritten text recognition is a problem rarely studied out of specific applications for which lexical knowledge can constrain the vocabulary to a limited one. In the case of handwritten text recognition, additional information can be exploited to characterize the specificity of the writing. This knowledge can help the recognition system to find coherent solutions from both the lexical and the morphological points of view. We present the principles of a handwritten text recognition system based on the online learning of the writer shapes. The proposed scheme is shown to improve the recognition rates on a sample of fifteen writings, unknown to the system.

Từ khóa

#Text recognition #Handwriting recognition #Writing #Hidden Markov models #Character recognition #Vocabulary #Shape #Image analysis #Pattern recognition #Semisupervised learning

Tài liệu tham khảo

10.1109/34.784288 dempster, 1977, Maximum likelihood from incomplete data via the em algorithm, Journal of the Royal Statistical Society, 39, 1 10.1016/S0031-3203(97)00023-X 10.1109/34.506792 10.1016/0031-3203(95)00072-0 lorette, 1992, Reconnaissance de textes manuscrits hors-ligne: Un proble?me d'analyse de sce?ne, Bigre 80 CNED'92, 109 10.1007/s100320050035 10.1016/S0167-8655(98)00039-7 10.1109/ICDAR.1999.791900 10.1109/ICDAR.1999.791885 10.1109/ICDAR.1999.791823