ARTIFICIAL INTELLIGENCE FOR THE AUTOMATIC REASSEMBLY AND VIRTUAL RECONSTRUCTION OF ARCHAEOLOGICAL POTTERY FRAGMENTS: A CRITICAL REVIEW OF METHODS, BENCHMARKS AND PROSPECTS FOR CENTRAL ASIAN ARCHAEOLOGY

Authors

  • Yusufjonov Fayoziddin Nodirjonovich Third-year student, Archaeology (Uzbek–Chinese Joint Archaeology Department), Faculty of History, Fergana State University, Fergana, Uzbekistan Author

Keywords:

archaeological ceramics; sherd reassembly; artificial intelligence; 3D scanning; virtual reconstruction; Fergana Valley.

Abstract

Archaeological pottery is usually recovered as thousands of broken, partly eroded sherds, and its manual reassembly is slow and requires physical handling of fragile material. This review examines how digitization and artificial intelligence (AI) are changing that task. Based on a structured, source-verified survey of peer-reviewed papers, conference proceedings and preprints, it traces the pipeline from 3D scanning through fracture-surface segmentation to pairwise and multi-fragment matching, and compares three method families: symmetry-guided geometric reassembly, learning-based pose estimation and generative completion. Results obtained on synthetic benchmarks and on small real datasets are not directly comparable: the highest reported part accuracy on the synthetic Breaking Bad benchmark is 90.2% (on a volume-constrained subset), whereas Structure-from-Sherds++ reached 87% reassembly accuracy on 142 real fragments of ten pots, and one batch-scanning system reports more than 700 sherds digitized per day. Erosion, missing fragments, mixed assemblages and the scarcity of real training data remain the main open problems. An expert-in-the-loop workflow is proposed, and its relevance for Uzbek–Chinese field projects in the Fergana Valley is discussed.

References

1.Lu, J., Han, H., Jiang, J., Liang, Y., Hua, J., Li, X., & Huang, Q. (2025). A survey on computational solutions for reconstructing complete objects by reassembling their fractured parts. arXiv preprint arXiv:2410.14770 (version of 28 March 2025). https://arxiv.org/abs/2410.14770

2.Karasik, A., & Smilansky, U. (2008). 3D scanning technology as a standard archaeological tool for pottery analysis: Practice and theory. Journal of Archaeological Science, 35(5), 1148–1168. https://doi.org/10.1016/j.jas.2007.08.008

3.Huang, Q.-X., Flöry, S., Gelfand, N., Hofer, M., & Pottmann, H. (2006). Reassembling fractured objects by geometric matching. ACM Transactions on Graphics, 25(3), 569–578. https://doi.org/10.1145/1141911.1141925

4.Brown, B. J., Toler-Franklin, C., Nehab, D., Burns, M., Dobkin, D., Vlachopoulos, A., Doumas, C., Rusinkiewicz, S., & Weyrich, T. (2008). A system for high-volume acquisition and matching of fresco fragments: Reassembling Theran wall paintings. ACM Transactions on Graphics (Proc. SIGGRAPH), 27(3). https://gfx.cs.princeton.edu/pubs/Brown_2008_ASF/fresco08_paper.pdf

5.Wang, J., Zhang, C., Wang, P., Li, X., Cobb, P. J., Theobalt, C., & Wang, W. (2023). Batch-based model registration for fast 3D sherd reconstruction. In Proceedings of the IEEE/CVF International Conference on Computer Vision (ICCV). https://doi.org/10.1109/ICCV51070.2023.01335. Open access: https://openaccess.thecvf.com/content/ICCV2023/html/Wang_Batch-based_Model_Registration_for_Fast_3D_Sherd_Reconstruction_ICCV_2023_paper.html

6.Yoo, S. J., Liu, S., Arshad, M. Z., et al. (2025). Structure-from-Sherds++: Robust incremental 3D reassembly of axially symmetric pots from unordered and mixed fragment collections. arXiv preprint arXiv:2502.13986. https://arxiv.org/abs/2502.13986

7.Sellán, S., Chen, Y.-C., Wu, Z., Garg, A., & Jacobson, A. (2022). Breaking Bad: A dataset for geometric fracture and reassembly. In Advances in Neural Information Processing Systems (NeurIPS), Datasets and Benchmarks Track. arXiv:2210.11463. https://doi.org/10.48550/arXiv.2210.11463

8.Navarro, P., Cintas, C., Lucena, M., Fuertes, J. M., et al. (2023). IberianVoxel: Automatic completion of Iberian ceramics for cultural heritage studies. In Proceedings of the Thirty-Second International Joint Conference on Artificial Intelligence (IJCAI-23). https://www.ijcai.org/proceedings/2023/647

9.Pinho, M., Mota, G., & Costa, G. (2025). A deep learning approach to assist in pottery reconstruction from its sherds [Preprint, not peer-reviewed]. Preprints.org, posted 17 February 2025. https://doi.org/10.20944/preprints202502.1174.v1

10.Papaioannou, G., Karabassi, E.-A., & Theoharis, T. (2001). Virtual archaeologist: Assembling the past. IEEE Computer Graphics and Applications, 21(2), 53–59. (Bibliographic data verified through the reference list of [3]; no open-access link verified.)

11.Wang, R. (2025, 8 May). Unearthing the secrets of a Central Asian crossroads. China Daily (Hong Kong edition). https://www.chinadailyhk.com/article/611248

12.Associated Press of Pakistan. (n.d.). Uzbekistan–China archaeologists discover ancient Kuva settlement [citing Xinhua News Agency]. Retrieved 10 November 2026, from https://www.app.com.pk/?p=1144542

13.Belt and Road archaeological cooperation fosters closer friendship among civilizations. (2024, March). Retrieved 10 November 2026, from https://kzxt.chinakongzi.org/English/CulturalHighlights/202403/t20240328_572168.htm

Downloads

Published

2026-11-11