Mo.Se.: Mosaic image segmentation based on deep cascading learning
Submitted: 2020-08-10
|Accepted: 2020-10-14
|Published: 2021-01-19
Downloads
Keywords:
cultural heritage, mosaic, deep learning, image segmentation, digitization
Supporting agencies:
Partially found within the framework of the project Innovative technologies and training activities for the conservation and enhancement of the archaeological site of Umm er-Rasas (Jordan) funded by Ministero degli Affari Esteri e della Cooperazione Inter
Abstract:
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.
References:
Bartoli, A., Fenu, G., Medvet, E., Pellegrino, F. A., & Timeus, N. (2016, November). Segmentation of Mosaic Images Based on Deformable Models Using Genetic Algorithms. In International Conference on Smart Objects and Technologies for Social Good (pp. 233-242). Springer, Cham. https://doi.org/10.1007/978-3-319-61949-1_25
Battiato, S., Di Blasi, G., Farinella, G. M., & Gallo, G. (2007, December). Digital mosaic frameworksâ€an overview. In computer graphics forum (Vol. 26, No. 4, pp. 794-812). Oxford, UK: Blackwell Publishing Ltd. https://doi.org/10.1111/j.1467-8659.2007.01021.x
Beucher, S., & Lantuéjoul, C. (1979). Use of watersheds in contour detection. International workshop on image processing: Real-time edge and motion detection/estimation. Rennes, France.
Benyoussef, L., & Derrode, S. (2011). Analysis of ancient mosaic images for dedicated applications. Digital Imaging for Cultural Heritage Preservation: Analysis, Restoration, and Reconstruction of Ancient Artworks, 385.
Bonfigli, R., Felicetti, A., Principi, E., Fagiani, M., Squartini, S., & Piazza, F. (2018). Denoising autoencoders for non-intrusive load monitoring: improvements and comparative evaluation. Energy and Buildings, 158. https://doi.org/10.1016/j.enbuild.2017.11.054
Bordoni, L., & Mele, F. (Eds.). (2016). Artificial intelligence for cultural heritage. Cambridge Scholars Publishing.
Bourke, P. (2014, December). Novel imaging of heritage objects and sites. In 2014 International Conference on Virtual Systems & Multimedia (VSMM) (pp. 25-30). IEEE. 10.1109/VSMM.2014.7136666
Çiçek, Ö., Abdulkadir, A., Lienkamp, S. S., Brox, T., & Ronneberger, O. (2016, October). 3D U-Net: learning dense volumetric segmentation from sparse annotation. In International conference on medical image computing and computer-assisted intervention (pp. 424-432). Springer, Cham. https://doi.org/10.1007/978-3-319-46723-8_49
Cipriani, L., & Fantini, F. (2017). Digitalization culture VS archaeological visualization: integration of pipelines and open issues. The International Archives of Photogrammetry, Remote Sensing and Spatial Information Sciences, 42, 195. https://doi.org/10.5194/isprs-archives-XLII-2-W3-195-2017
Djibril, M. O., & Thami, R. O. H. (2008). Islamic geometrical patterns indexing and classification using discrete symmetry groups. Journal on Computing and Cultural Heritage (JOCCH), 1(2), 1-14. https://doi.org/10.1145/1434763.1434767
Djibril, M. O., Thami, R. O. H., Benslimane, R., & Daoudi, M. (2005). Une nouvelle technique pour l'indexation des arabesques basée sur la dimension fractale. Univ. Mohamed V, Maroc.
Falk, T., Mai, D., Bensch, R., Çiçek, Ö., Abdulkadir, A., Marrakchi, Y., Böhm, A., Deubner, J., Jäckel, Z., Seiwald, K., & Dovzhenko, A. (2019). U-Net: deep learning for cell counting, detection, and morphometry. Nature methods, 16(1), 67-70. https://doi.org/10.1038/s41592-018-0261-2
Felicetti, A., Albiero, A., Gabrielli, R., Pierdicca, R., Paolanti, M., Zingaretti, P., & Malinverni, E. S. (2018). Automatic Mosaic Digitalization: a Deep Learning approach to tessera segmentation. In METROARCHEO, IEEE International Conference on Metrology for Archaeology and Cultural Heritage. Cassino. https://doi.org/10.1109/MetroArchaeo43810.2018.13606
Fenu, G., Jain, N., Medvet, E., Pellegrino, F. A., & Namer, M. P. (2015, March). On the Assessment of Segmentation Methods for Images of Mosaics. In VISAPP (3) (pp. 130-137). https://doi.org/10.13140/RG.2.1.3025.6489
Fenu, G., Medvet, E., Panfilo, D., & Pellegrino, F. A. (2020). Mosaic Images Segmentation using U-net. In International Conference on Pattern Recognition Applications and Methods (pp. 485-492). Scitepress. http://dx.doi.org/10.5220/0008967404850492
Fontanella, F., Molinara, M., Gallozzi, A., Cigola, M., Senatore, L. J., Florio, R., Clini, P., & Celis, F. (2019, June). HeritageGO (HeGO) A Social Media Based Project for Cultural Heritage Valorization. In Adjunct Publication of the 27th Conference on User Modeling, Adaptation and Personalization (pp. 377-382). https://doi.org/10.1145/3314183.3323863
Gil, F. A., Gomis, J. M., & Pérez, M. (2009). Reconstruction Techniques for Image Analysis of Ancient Islamic Mosaics. International Journal of Virtual Reality, 8(3), 5-12. https://doi.org/10.20870/IJVR.2009.8.3.2735
Kingma, D. P., & Ba, J. (2014). Adam: A method for stochastic optimization. arXiv preprint arXiv:1412.6980.
Kohl, S., Romera-Paredes, B., Meyer, C., De Fauw, J., Ledsam, J. R., Maier-Hein, K., Eslami, S.M.A, Rezende, D.J., & Ronneberger, O. (2018). A probabilistic u-net for segmentation of ambiguous images. In Advances in Neural Information Processing Systems (pp. 6965-6975). https://arxiv.org/abs/1806.05034
Liciotti, D., Paolanti, M., Pietrini, R., Frontoni, E., & Zingaretti, P. (2018, August). Convolutional networks for semantic heads segmentation using top-view depth data in crowded environment. In 2018 24th international conference on pattern recognition (ICPR) IEEE. https://doi.org/10.1109/ICPR.2018.8545397
Maghrebi, W., Ammar, A. B., Alimi, A. M., & Khabou, M. A. (2013). An Intelligent mutli-object retrieval system for historical mosaics. Editorial Preface, 4(4). https://doi.org/10.14569/IJACSA.2013.040417
Maghrebi, W., Baccour, L., Khabou, M. A., & Alimi, A. M. (2007, November). An indexing and retrieval system of historic art images based on fuzzy shape similarity. In Mexican International Conference on Artificial Intelligence (pp. 623-633). Springer, Berlin, Heidelberg. https://doi.org/10.1007/978-3-540-76631-5_59
Maghrebi, W., Borchani, A., Khabou, M. A., & Alimi, A. M. (2007, September). A system for historic document image indexing and retrieval based on xml database conforming to mpeg7 standard. In International Workshop on Graphics Recognition (pp. 114-125). Springer, Berlin, Heidelberg. https://doi.org/10.1007/978-3-540-88188-9_12
Malinverni, E. S., Pierdicca, R., Di Stefano, F., Gabrielli, R., & Albiero, A. (2019). Virtual museum enriched by GIS data to share science and culture. Church of Saint Stephen in Umm Ar-Rasas (Jordan). Virtual Archaeology Review, 10(21). https://doi.org/10.4995/var.2019.11919
M'hedhbi, M., Mezhoud, R., M'hiri, S., & Ghorbel, F. (2006, April). A new content-based image indexing and retrieval system of mosaic images. In 2006 2nd International Conference on Information & Communication Technologies (Vol. 1, pp. 1715-1719). IEEE. https://doi.org/10.1109/ICTTA.2006.1684644
Pierdicca, R., Frontoni, E., Malinverni, E. S., Colosi, F., & Orazi, R. (2016). Virtual reconstruction of archaeological heritage using a combination of photogrammetric techniques: Huaca Arco Iris, Chan Chan, Peru. Digital Applications in Archaeology and Cultural Heritage, 3(3). https://doi.org/10.1016/j.daach.2016.06.002
Pierdicca, R., Frontoni, E., Zingaretti, P., Malinverni, E. S., Colosi, F., & Orazi, R. (2015, August). Making visible the invisible. Augmented reality visualization for 3D reconstructions of archaeological sites. In International Conference on Augmented and Virtual Reality (Blinded for peer review). Springer, Cham. https://doi.org/10.1007/978-3-319-22888-4_3
Ronneberger, O., Fischer, P., & Brox, T. (2015, October). U-net: Convolutional networks for biomedical image segmentation. In International Conference on Medical image computing and computer-assisted intervention (pp. 234-241). Springer, Cham. https://doi.org/10.1007/978-3-319-24574-4_28
Vincent, L., & Soille, P. (1991). Watersheds in digital spaces: an efficient algorithm based on immersion simulations. IEEE Transactions on Pattern Analysis & Machine Intelligence, (6), 583-598. https://doi.org/10.1109/34.87344
Youssef, L. B., & Derrode, S. (2008). Tessella-oriented segmentation and guidelines estimation of ancient mosaic images. Journal of Electronic Imaging, 17(4), 043014. https://doi.org/10.1117/1.3013543
Zarghili, A., Gadi, N., Benslimane, R., & Bouatouch, K. (2001). Arabo-Moresque decor image retrieval system based on mosaic representations. Journal of Cultural Heritage, 2(2), 149-154. https://doi.org/10.1016/S1296-2074(01)01116-5
Zarghili, A., Kharroubi, J., & Benslimane, R. (2008). Arabo-Moresque decor images retrieval system based on spatial relationships indexing. Journal of cultural heritage, 9(3), 317-325. https://doi.org/10.1016/j.culher.2007.10.008
Zitová, B., Flusser, J., & Šroubek, F. (2004). An application of image processing in the medieval mosaic conservation. Pattern analysis and applications, 7(1), 18-25. https://doi.org/10.1007/s10044-003-0200-3
