Digital techniques for character recognition in old documents

Two research projects have recently been activated by the Italian National Research Council (CNR) with the objective of designing and developing computerised tools to retrieve and restore textual information contained in ancient documents, accessed as digital images. The first, LAperLA, concerns old printed books, while the second, carried out in the framework of a CNR special project that aims at preserving our cultural heritage, regards old manuscripts. Both projects are aimed at implementing an integrated system that improves the quality of the images and, at the same time, activates optical character recognition (OCR) functions. Three CNR Institutes are involved in this activity: the Istituto di Linguistica Computazionale (ILC-Pisa), the Istituto di Elaborazione della Informazione (IEI-Pisa) and the Istituto di Applicazioni del Calcolo (IAC-Rome).