OCR / HTR TECHNOLOGIES AND ARMENIAN HERITAGE PRESERVATION*

Abstract

How OCR and HTR Technologies for the Armenian Language Can Help Preserve Heritage OCR (Optical Character Recognition) and HTR (Handwritten Text Recognition) are now available for the Armenian language. This technology can offer greater valorization for documents by enabling improved accessibility and findability via keyword search, but also presents a new challenge for digital libraries. Our talk intends to present the modern challenges that arose during the process of Armenian text recognition, as well as show the modern possibilities. A focus will be made on the technology developed by Calfa for handwritten archives, ancient manuscripts and old printed books. Feedback on three of our ongoing projects: processing of the catalogue of Armenian manuscripts (Mekhitarist, Venice), newspapers of FSL of NASRA, and Armenian letters (Mekhitarist, Venice) are presented.Methodology applied by Calfa leads to an accuracy greater than 98% for handwritten documents and greater than 99.9% for printed documents.