When Tesseract Meets PERO : Open-Source Optical Character Recognition of Medieval Texts

Warning

This publication doesn't include Faculty of Economics and Administration. It includes Faculty of Informatics. Official publication website can be found on muni.cz.
Authors

NOVOTNÝ Vít HORÁK Aleš

Year of publication 2022
Type Article in Proceedings
Conference Proceedings of the Sixteenth Workshop on Recent Advances in Slavonic Natural Languages Processing, RASLAN 2022.
MU Faculty or unit

Faculty of Informatics

Citation
Web
Keywords optical character recognition; OCR; medieval texts; AHISTO project
Description Conversion of scanned images to the text form, denoted as optical character recognition or OCR, for contemporary printed texts is widely considered a solved problem. However, the optical character recognition of early printed books and reprints of medieval texts remains an open challenge. In our previous work, we developed an end-to-end image-to-text pipeline (via optical character recognition) for medieval texts, named AHISTO OCR, and we released it together with our test dataset under open licenses. However, the published system relied on the closed-source Google Vision AI service as one component, which made the experiments less reproducible. In this work, we replace Google Vision AI with an open-source OCR algorithm named PERO and we show that this not only makes the AHISTO OCR pipeline open, but also improves the performance of the system. We release the updated AHISTO OCR system and its test results again under open licenses.
Related projects:

You are running an old browser version. We recommend updating your browser to its latest version.