Creating a dataset of digital images of Byzantine frescoes for analysis and classification using machine learning techniques
International Symposium on the Conservation of Monuments in the Mediterranean Basin · 19.07.2024 · 177-182
DOI: 10.12681/monubasin.8183 ↗Kurz gefasst
Fachartikel in der Zeitschrift International Symposium on the Conservation of Monuments in the Mediterranean Basin, online erschienen am 19.07.2024. Verfasst von Antigoni Vlisidi und 1 weiteren Autor:innen. Die Arbeit nennt 15 Literaturangaben.
Abstract (Originalsprache)
The methodology for creating a database of fresco images, mainly of the Macedonian School (13th-14th century), is presented here. This base will be used later for digital image processing for classification using machine learning. For this purpose, modern digital data analysis tools in Python were used. The collection of appropriate material, the digitization of images, the input of information and the categories examined are presented in detail, along with the tools used. One of the main problems of the research was that there were no open digital libraries to cover with images of the specific period examined, so the material gathered comes from corporate sources such as books, studies and collectors. After analyzing the art historical information from Byzantine frescoes not only for their measurability and reliability, but also for availability, a python application that manages the data as an object-oriented relational graph is created. This is a first approach to such material that combines easy-to-use tools with historical knowledge and provides a ready-to-use and development base.
Zitieren
APA
Vlisidi, A., & Moropoulou, A. (2024). Creating a dataset of digital images of Byzantine frescoes for analysis and classification using machine learning techniques. International Symposium on the Conservation of Monuments in the Mediterranean Basin, 177-182. https://doi.org/10.12681/monubasin.8183
Vancouver
Vlisidi A, Moropoulou A. Creating a dataset of digital images of Byzantine frescoes for analysis and classification using machine learning techniques. International Symposium on the Conservation of Monuments in the Mediterranean Basin. 2024:177-182. doi:10.12681/monubasin.8183
BibTeX
@article{Vlisidi2024,
author = {Vlisidi, Antigoni and Moropoulou, Antonia},
title = {Creating a dataset of digital images of Byzantine frescoes for analysis and classification using machine learning techniques},
journal = {International Symposium on the Conservation of Monuments in the Mediterranean Basin},
year = {2024},
pages = {177--182},
doi = {10.12681/monubasin.8183}
}
EUVIMED