History
This project builds on From Papyrus to Pixels: Optical Character Recognition Applied to Ancient Egyptian Hieratic by Julius Tabin (2022), with code and data on GitHub. The web application was built by Mark-Jan Nederhof, in collaboration with Julius Tabin and Christian Casey.
Aims
This project aims to digitize facsimiles of different hieratic texts, annotated with shapes of glyphs. Image processing techniques are used for automatic segmentation, and OCR techniques are used for automatic classification.
Given the database of shapes, techniques of dimensionality reduction can used to compare manuscripts from different periods, provenances and genres, and to identify different realizations of the same underlying character.
Hosting
This project is hosted in partnership with the Egyptology department of the University of Liège.
The source code is on GitHub.
Bibliography
J. Tabin, M.-J. Nederhof, and C. Casey. Collaborative Annotation and Computational Analysis of Hieratic. In M. Coustaty and A. Fornés (eds.), Document Analysis and Recognition -- ICDAR 2023 Workshops, Lecture Notes in Computer Science, volume 14193, Part 1, pages 267-283, San José, CA, USA, 2023. Springer-Verlag.
Text Provenance Genre Period Creator Ebers unknown medical 18 dyn Möller Hymn to Senwosret III Lahun hymn 12 dyn Möller Lahun Temple Files Lahun administrative 12 dyn Möller Peasant B1 Thebes literary 12 dyn Tabin Peasant R Thebes literary 13 dyn Möller Prisse Thebes instruction 12 dyn Möller Rhind unknown mathematical 15 dyn Möller Shipwrecked unknown literary 12 dyn Tabin Sinuhe B Thebes literary 12 dyn Möller Sinuhe R Thebes literary 13 dyn Möller Texte aus Hatnub Hatnub graffiti 12 dyn Möller Westcar unknown literary 18 dyn Möller Will of Wah Lahun administrative 12 dyn Möller
No comments:
Post a Comment