Automatic Text Recognition Made Easy
Textual sources, primarily handwritten, and found in archives or libraries, are critical for humanities research. These sources, whether scanned by researchers or digitized by cultural institutions, need to be accessible in full text to facilitate (computational) analysis.
This project leverages Automatic Text Recognition (ATR) technologies to transform extensive text corpora into searchable transcriptions. We have developed a series of video and text tutorials that detail the application of ATR technology, providing step-by-step guidance through the typical ATR workflow and explaining main concepts and terms.
Our aim is to equip researchers, librarians, and archivists with the knowledge and tools to efficiently convert images of textual heritage materials into editable text, enhancing their ability to engage with and utilise these resources effectively.
The tutorials, available in English with subtitles in English, German, and French, are designed to be concise and accessible to scholars who may not be prepared to immediately engage with and apply complex technical documentation.
This blog accompanies the project, called “Harmonising Workflows in ATR. Publication Pipelines of Textual Heritage”. It is funded by DARIAH-EU and led by Anne Baillot and Mareike König.
OpenEdition suggests that you cite this post as follows:
Mareike König (February 16, 2023). Automatic Text Recognition Made Easy. Automatic Text Recognition. Retrieved April 22, 2025 from https://doi.org/10.58079/11npt