Libraries and cultural organisations have a rich amount of digitised historical handwritten material. A vast majority of this material has not yet been transcribed. The task of making these historical collections available for public access is challenging, especially in performing a simple text search across the collection.
Machine learning based methods for handwritten text recognition are gaining importance these days, which require huge amount of pre- transcribed texts for training the system. However, a vast amount of documents are not yet transcribed.
Therefore, this paper presents a training- free word spotting algorithm that can be used as a search tool, where case studies on Alvin (Swedish repository) and Clavius on the Web are presented.
The image shows how a user has searched for the word ‘Berlin’ in the diary of the famous Swedish author Karin Boye. The found words are marked with a blue bounding box.
-
Making Large Collections of Handwritten Material Easily Accessible and Searchable
A. Hast, P. Cullhed, E. Vats, M. Abrate.
Proceedings of the Italian Research Conference on Digital Libraries (IRCDL), Full Paper. pp. 1-11. 2019.
@inproceedings{Hast1278821,
author = {Hast, Anders and Cullhed, Per and Vats, Ekta and Abrate, Matteo},
booktitle = {Digital Libraries : Supporting Open Science},
institution = {Uppsala University, Division of Visual Information and Interaction},
institution = {Uppsala University, Computerized Image Analysis and Human-Computer Interaction},
institution = {Uppsala University, University Library},
pages = {18--28},
title = {Making large collections of handwritten material easily accessible and searchable},
series = {Communications in Computer and Information Science},
number = {988},
DOI = {10.1007/978-3-030-11226-4_2},
ISBN = {978-3-030-11225-7},
year = {2019}
}
Anders Hast 