Abstract

ABSTRACTWithin the framework of the project aiming at the automatic processing of post cheques inFrench Post Financial establishments, the French Post Technical Research Department (SRTP) has started studying the automatic reading of cursive script amounts on cheque. We present the total processing sequence and in particular the part related to the retrieval ofscript information on the cheque. Indeed, pre-printed information (guiding lines, slashes, acronyms,typed information) and script information, which is the one to be identified, coexist on the cheque. The problem consists in retrieving the information to be processed from the cheque without damaging the information. 1. INTRODUCTIONOne of the most signifiant problems in Optical Character Recognition (OCR) is the conversionof non ideal images into ideal binary images whitout degradation of the informations located on the document . Indeed the information that needs to be processed is very often accompagnied by patterns and backgrounds of pictures, depending on the paper support used. This information must

Full Text
Published version (Free)

Talk to us

Join us for a 30 min session where you can share your feedback and ask us any queries you have

Schedule a call