Handwritten Text Recognition (HTR)

Our automatic handwritten text recognition (HTR) service is exclusively available within our PastView management and publishing platform. Utilising AI, our Handwritten Text Recognition service can recognise a huge array of handwriting styles from historical documents, automatically adding this information as metadata to support enhanced searching of digitised collections.

Handwritten Text Recognition is invaluable in making the information within digitised handwritten journals, diaries, letters or even notes beneath photographs more accessible. Providing the most efficient and instantaneous retrieval of your valuable content, TownsWeb Archiving are proud to announce that our HTR service is made available to all users of our PastView platform.

Frequently Asked Questions – Find out more

Highly accurate data capture

Our HTR service accurately extracts a huge array of handwriting styles from historical documents, automatically adding this information as metadata for your digitised materials.

Capture data from handwritten material

We can capture written text from any of your digitised material, including handwritten notes, personal diaries, and even text within photographs.

Instantly searchable digital content

HTR results are available directly within the PastView platform. Creating fully searchable metadata, revolutionising the access and discovery of your handwritten collections.

Handwritten Text-Recognition (HTR) comes to PastView!

Discover how PastView is revolutionising digital collections with Handwritten Text Recognition (HTR), enhancing searching capabilities.

Learn more about HTR

Handwritten Text Recognition (HTR) Frequently Asked Questions

Our cutting edge Handwritten Text Recognition (HTR) service scans your JPEG image collections (often produced via digitisation) during or after upload to the PastView platform. Our HTR service automatically recognises handwritten text within the images, and converts this text into machine-readable digital text documents.

The recognised text is then added as metadata to your items withi PastView and associated with the image it was scanned from, to allow keyword searching of the text content.

Our Handwritten Text Recognition (HTR) service can automatically recognise handwritten text within digitised images (JPEGs), recognising both modern and historic writing styles, including historic letters and spellings.

Performing Handwritten Text Recognition (HTR) on modern and historic writing styles can produce very accurate results.

For cursive handwriting and older lettering styles, you may find better accuracy by using our specialist heritage HTR service which has been specifically trained on heritage documents. However, we encourage all customers to try both first as this improved recognition is not guaranteed.

It should be noted that damages to the original material can also negatively impact the results.

Our Handwritten Text Recognition (HTR) service is suitable for many handwritten materials in your archive. There are some cases, however, where handwritten transcription or data entry conducted by specialist technicians may be more suitable e.g When handwritting is of low legibility, the original material is damaged or the handwritten text is organised into a structured format such as a register.