ARTICLE | doi:10.20944/preprints201909.0062.v1
Online: 5 September 2019 (12:19:48 CEST)
This paper presents a simple yet effective solution for the transcrip- tion of printed texts. Our tool consists of a web-based user interface that provides an easy-to-use and ergonomic workflow and a col- laborative environment for the philologists while allowing them to profit from machine learning OCR technology. As the targeted use case is not mass digitisation but the creation of accurate citable digital editions, the user interface for ground truth production and post correction is built to provide the means for rapid proofread- ing while minimising the amount of errors. The productivity of the setup is further improved by enabling progressive OCR train- ing and recognition in the background to constantly increase the accuracy of the predictions. The advantages of the application are showcased in the second part of the paper by documenting our experiences utilising it for di- gitising Arabic and Latin texts. Over the course of several months the tool has been used to create transcriptions of a wide range of sources, among them challenging early modern editions and Ar- abic scripts, producing a large amount of reusable OCR training data as a positive side effect. Finally, there will be a discussion of possible future extensions of the tool and of how it could be adapted to fit the needs of other digitisation projects.
ARTICLE | doi:10.20944/preprints201909.0101.v1
Subject: Mathematics & Computer Science, Artificial Intelligence & Robotics Keywords: Optical Character Recognition; Document Analysis; Historical Printings
Online: 9 September 2019 (12:08:16 CEST)
Optical Character Recognition (OCR) on historical printings is a challenging task mainly due to the complexity of the layout and the highly variant typography. Nevertheless, in the last few years great progress has been made in the area of historical OCR, resulting in several powerful open-source tools for preprocessing, layout recognition and segmentation, character recognition and post-processing. The drawback of these tools often is their limited applicability by non-technical users like humanist scholars and in particular the combined use of several tools in a workflow. In this paper we present an open-source OCR software called OCR4all, which combines state-of-the-art OCR components and continuous model training into a comprehensive workflow. A comfortable GUI allows error corrections not only in the final output, but already in early stages to minimize error propagations. Further on, extensive configuration capabilities are provided to set the degree of automation of the workflow and to make adaptations to the carefully selected default parameters for specific printings, if necessary. Experiments showed that users with minimal or no experience were able to capture the text of even the earliest printed books with manageable effort and great quality, achieving excellent character error rates (CERs) below 0.5%. The fully automated application on 19th century novels showed that OCR4all can considerably outperform the commercial state-of-the-art tool ABBYY Finereader on moderate layouts if suitably pretrained mixed OCR models are available. The architecture of OCR4all allows the easy integration (or substitution) of newly developed tools for its main components by standardized interfaces like PageXML, thus aiming at continual higher automation for historical printings.