Handwritten Text Recognition Through Writer Adaptation
IWFHR '02 Proceedings of the Eighth International Workshop on Frontiers in Handwriting Recognition (IWFHR'02)
DAS'06 Proceedings of the 7th international conference on Document Analysis Systems
Hi-index | 0.00 |
In this communication, we propose a method for theautomatic extraction of numerical fields in handwrittendocuments. The approach exploits the known syntacticstructure of the numerical field to extract, combined witha set of contextual morphological features to find the bestlabel to each connected component. Applying an HMMbased syntactic analyzer on the overall document allowsto localize/extract fields of interest. Reported results onthe extraction of zip codes, phone numbers and customercodes from handwritten incoming mail documentsdemonstrate the interest of the proposed approach.