Dariah.lab
LOGO UE
LOGO UE

HTR: Technology in the service of history

HTR (Handwritten Text Recognition) from materials such as scans and photos opens new possibilities for researchers of the past, facilitating the analysis of historical sources.

Using advanced OCR (Optical Character Recognition) and HTR technologies, old prints and manuscripts can be transformed into electronic text form. Thanks to this, documents become more accessible and searchable, and can also be used as a basis for creating their digital edition. Modern handwriting recognition methods are based on machine learning techniques, accelerating and improving the quality of the manuscript analysis process.


Dariah HTR: transcription tools

Dariah HTR is a service based on eScriptorium software and the Kraken system. These open source tools provide a comprehensive work environment for historical manuscripts and prints, enabling manual and automatic transcription of handwritten scans.

  • eScriptorium: web application for working on historical texts, enabling the creation and management of scan collections, support for the IIIF protocol and supporting collaboration between users.
  • Model training and automatic transcription: eScriptorium uses modern machine learning methods, numerous ready-made recognition models tailored to different types of writing created by the research community, and allows you to train your own models better suited to specific manuscript collections.
  • Unlimited data export: transcription can be saved in standard formats such as Alto, PAGE XML or TXT, trained models can also be exported from the application or shared with other eScriptorium users.

The Dariah HTR service is available for a fee on the servers of the Poznan Supercomputing and Networking Center: the link will be published soon.


The Digital History Lab of the IH PAN offers services in the field of:

  • preparing the environment for transcription: adapting and importing scans
  • help in training models based on training material
  • post-processing and consultation to improve transcription results
  • training and support in the use of eScriptorium.

Contact us to learn more about how we can support your research with handwriting recognition technology. We provide professional support at every stage of working with historical text materials.

The provision of research services is subject to payment.