Mo.Se.: Mosaic image segmentation based on deep cascading learning

Mosaic is an ancient type of art used to create decorative images or patterns combining small components. A digital version of a mosaic can be useful for archaeologists, scholars and restorers who are interested in studying, comparing and preserving mosaics. Nowadays, archaeologists base their studi...

Full description

Saved in:
Bibliographic Details
Main Authors: Andrea Felicetti, Marina Paolanti, Primo Zingaretti, Roberto Pierdicca, Eva Savina Malinverni
Format: Article
Language:English
Published: Universitat Politècnica de València 2021-01-01
Series:Virtual Archaeology Review
Subjects:
Online Access:https://polipapers.upv.es/index.php/var/article/view/14179
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1850058694779207680
author Andrea Felicetti
Marina Paolanti
Primo Zingaretti
Roberto Pierdicca
Eva Savina Malinverni
author_facet Andrea Felicetti
Marina Paolanti
Primo Zingaretti
Roberto Pierdicca
Eva Savina Malinverni
author_sort Andrea Felicetti
collection DOAJ
description Mosaic is an ancient type of art used to create decorative images or patterns combining small components. A digital version of a mosaic can be useful for archaeologists, scholars and restorers who are interested in studying, comparing and preserving mosaics. Nowadays, archaeologists base their studies mainly on manual operation and visual observation that, although still fundamental, should be supported by an automatized procedure of information extraction. In this context, this research explains improvements which can change the manual and time-consuming procedure of mosaic tesserae drawing. More specifically, this paper analyses the advantages of using Mo.Se. (Mosaic Segmentation), an algorithm that exploits deep learning and image segmentation techniques; the methodology combines U-Net 3 Network with the Watershed algorithm. The final purpose is to define a workflow which establishes the steps to perform a robust segmentation and obtain a digital (vector) representation of a mosaic. The detailed approach is presented, and theoretical justifications are provided, building various connections with other models, thus making the workflow both theoretically valuable and practically scalable for medium or large datasets. The automatic segmentation process was tested with the high-resolution orthoimage of an ancient mosaic by following a close-range photogrammetry procedure. Our approach has been tested in the pavement of St. Stephen's Church in Umm ar-Rasas, a Jordan archaeological site, located 30 km southeast of the city of Madaba (Jordan). Experimental results show that this generalized framework yields good performances, obtaining higher accuracy compared with other state-of-the-art approaches. Mo.Se. has been validated using publicly available datasets as a benchmark, demonstrating that the combination of learning-based methods with procedural ones enhances segmentation performance in terms of overall accuracy, which is almost 10% higher. This study’s ambitious aim is to provide archaeologists with a tool which accelerates their work of automatically extracting ancient geometric mosaics. Highlights: • A Mo.Se. (Mosaic Segmentation) algorithm is described with the purpose to perform robust image segmentation to automatically detect tesserae in ancient mosaics. • This research aims to overcome manual and time-consuming procedure of tesserae segmentation by proposing an approach that uses deep learning and image processing techniques, obtaining a digital replica of a mosaic. • Extensive experiments show that the proposed framework outperforms state-of-the-art methods with higher accuracy, even compared with publicly available datasets.
format Article
id doaj-art-8ce70334191d4c6f9bb4dc954aabd1b1
institution DOAJ
issn 1989-9947
language English
publishDate 2021-01-01
publisher Universitat Politècnica de València
record_format Article
series Virtual Archaeology Review
spelling doaj-art-8ce70334191d4c6f9bb4dc954aabd1b12025-08-20T02:51:04ZengUniversitat Politècnica de ValènciaVirtual Archaeology Review1989-99472021-01-011224253810.4995/var.2021.141798387Mo.Se.: Mosaic image segmentation based on deep cascading learningAndrea Felicetti0Marina Paolanti1Primo Zingaretti2Roberto Pierdicca3Eva Savina Malinverni4Università Politecnica delle MarcheUniversità Politecnica delle MarcheUniversità Politecnica delle MarcheUniversitá Politecnica delle MarcheUniversità Politecnica delle MarcheMosaic is an ancient type of art used to create decorative images or patterns combining small components. A digital version of a mosaic can be useful for archaeologists, scholars and restorers who are interested in studying, comparing and preserving mosaics. Nowadays, archaeologists base their studies mainly on manual operation and visual observation that, although still fundamental, should be supported by an automatized procedure of information extraction. In this context, this research explains improvements which can change the manual and time-consuming procedure of mosaic tesserae drawing. More specifically, this paper analyses the advantages of using Mo.Se. (Mosaic Segmentation), an algorithm that exploits deep learning and image segmentation techniques; the methodology combines U-Net 3 Network with the Watershed algorithm. The final purpose is to define a workflow which establishes the steps to perform a robust segmentation and obtain a digital (vector) representation of a mosaic. The detailed approach is presented, and theoretical justifications are provided, building various connections with other models, thus making the workflow both theoretically valuable and practically scalable for medium or large datasets. The automatic segmentation process was tested with the high-resolution orthoimage of an ancient mosaic by following a close-range photogrammetry procedure. Our approach has been tested in the pavement of St. Stephen's Church in Umm ar-Rasas, a Jordan archaeological site, located 30 km southeast of the city of Madaba (Jordan). Experimental results show that this generalized framework yields good performances, obtaining higher accuracy compared with other state-of-the-art approaches. Mo.Se. has been validated using publicly available datasets as a benchmark, demonstrating that the combination of learning-based methods with procedural ones enhances segmentation performance in terms of overall accuracy, which is almost 10% higher. This study’s ambitious aim is to provide archaeologists with a tool which accelerates their work of automatically extracting ancient geometric mosaics. Highlights: • A Mo.Se. (Mosaic Segmentation) algorithm is described with the purpose to perform robust image segmentation to automatically detect tesserae in ancient mosaics. • This research aims to overcome manual and time-consuming procedure of tesserae segmentation by proposing an approach that uses deep learning and image processing techniques, obtaining a digital replica of a mosaic. • Extensive experiments show that the proposed framework outperforms state-of-the-art methods with higher accuracy, even compared with publicly available datasets.https://polipapers.upv.es/index.php/var/article/view/14179cultural heritagemosaicdeep learningimage segmentationdigitization
spellingShingle Andrea Felicetti
Marina Paolanti
Primo Zingaretti
Roberto Pierdicca
Eva Savina Malinverni
Mo.Se.: Mosaic image segmentation based on deep cascading learning
Virtual Archaeology Review
cultural heritage
mosaic
deep learning
image segmentation
digitization
title Mo.Se.: Mosaic image segmentation based on deep cascading learning
title_full Mo.Se.: Mosaic image segmentation based on deep cascading learning
title_fullStr Mo.Se.: Mosaic image segmentation based on deep cascading learning
title_full_unstemmed Mo.Se.: Mosaic image segmentation based on deep cascading learning
title_short Mo.Se.: Mosaic image segmentation based on deep cascading learning
title_sort mo se mosaic image segmentation based on deep cascading learning
topic cultural heritage
mosaic
deep learning
image segmentation
digitization
url https://polipapers.upv.es/index.php/var/article/view/14179
work_keys_str_mv AT andreafelicetti mosemosaicimagesegmentationbasedondeepcascadinglearning
AT marinapaolanti mosemosaicimagesegmentationbasedondeepcascadinglearning
AT primozingaretti mosemosaicimagesegmentationbasedondeepcascadinglearning
AT robertopierdicca mosemosaicimagesegmentationbasedondeepcascadinglearning
AT evasavinamalinverni mosemosaicimagesegmentationbasedondeepcascadinglearning