AI-ASSISTED DIGITALISATION OF HISTORICAL DOCUMENTS

Abstract. Preserving historical archival heritage involves not only physical measures to safeguard these valuable texts but also providing for their digital preservation. However, merely digitising manuscripts and codexes is enough. A further step needed: the digitalisation of content, i.e. verbatim transcription scanned texts. This process enables accurate preservation textual making it easier search information conduct analyses. With help artificial intelligence, particularly Deep Neural Networks (DNNs), automatic handwriting recognition can be performed. In this study, we employed a Convolutional Recurrent Network (CRNN), an established type DNN, determine minimum amount labelled data required automatically transcribe five different datasets that vary in language time period. The results show Character Error Rate (CER) lower than 10% achieved with just few hundred text lines almost all cases.