TexRec – Virtual Reconstruction, Interpretation and Preservation of the Textile Artefacts from the Oseberg Find

• Khawaja, Arsalan; Gigilashvili, Davit; Łojewski, Tomasz; George, Sony; Marzani, Franck & Hardeberg, Jon Yngve [Show all 7 contributors for this article] (2026). Beyond relighting: RTI for clustering fragmented heritage textiles using deep learning. npj Heritage Science. 14, p. 1–24. doi: 10.1038/s40494-026-02326-9. Full text in Research Archive Show summary

  This paper introduces a computational framework for clustering and visualizing textile fragment relationships using Reflectance Transformation Imaging (RTI). Our approach leverages deep learning models and utilizes the Hemispherical Harmonics (HSH) to extract discriminative features. The RTI data is first modeled using HSH before feature extraction by the deep feature extractor. The feature vectors are visualized through dimensionality reduction techniques, which help reveal the relationship between fragments by clustering them. We tested the proposed framework on the Oseberg textile collection (an open archeological artifact assembly problem) and a control dataset of Polish Dragoons textiles, demonstrating that the algorithm achieves good intra-class similarity and inter-class separation, distinguishing different textiles. RTI-based framework achieves higher clustering and dimensionality reduction evaluation scores between related fragments than RGB photography. The results confirm RTI’s potential as a data-rich, non-destructive imaging technique for supporting archeological reconstruction.

• Fadillah, Riestiya Zain; Gigilashvili, Davit; Havgar, Margrethe; Vedeler, Marianne & Hardeberg, Jon Yngve (2026). Automated thread counting in archaeological textiles using pulse-based feature extraction. The European Physical Journal Plus. 141, p. 1–28. doi: 10.1140/epjp/s13360-025-07194-z. Full text in Research Archive Show summary

  Abstract Archaeological textiles are often highly fragmented, posing significant challenges for reconstruction and analysis. Human experts traditionally rely on technical analyses, such as thread counting, to match fragments and determine their origins. However, this process is labour-intensive and time-consuming, especially for degraded artefacts such as the Oseberg tapestry from the Viking Age. Despite the potential for automation, existing computational methods have proven insufficiently robust, lacking the nuanced approach that human expertise offers. We propose a framework for automating thread counting in textile images using a feature extractor to capture thread counts in vertical and horizontal directions. Our method models thread counting as a pulse counting problem, inspired by how humans intuitively count threads. It involves analysing traces of pixel values from binary and gradient images to estimate thread density. Our method was compared to measurements by human experts, and while it tended to overestimate thread counts, it showed promising results in clustering fragments. Specifically, 14 out of 27 fragments were grouped in alignment with archaeologists’ hypotheses regarding their origins. Our method offers a valuable tool for digitally reconstructing archaeological textiles, improving efficiency in analysing and matching fragmented artefacts. Future work will focus on refining our approach with advanced thresholding techniques, modelling complex weaving techniques and utilizing hyperspectral images to improve accuracy in identifying complex weaving structures and mixed thread materials.

• Khawaja, Arsalan; Gigilashvili, Davit; Łojewski, Tomasz; George, Sony; Marzani, Franck & Hardeberg, Jon Yngve [Show all 7 contributors for this article] (2025). RTI beyond relighting: Image analysis for archaeological Oseberg textiles. Journal of Physics: Conference Series (JPCS). ISSN 1742-6588. 3128(1), p. 1–8. doi: 10.1088/1742-6596/3128/1/012017. Full text in Research Archive Show summary

  Abstract Reflectance Transformation Imaging (RTI), a non-destructive imaging technique, captures data rich in geometric and appearance information. However, it has traditionally been limited to relighting. We demonstrate RTI’s expanded potential through analysis of the Oseberg textiles, degraded and fragmented archaeological artifacts that present analysis challenges under conventional imaging methods. We present feature maps by applying statistical and geometric analysis to RTI data and a visualization tool. The computational analysis of RTI data can reveal textile characteristics that remain obscured under traditional photography.

• Gigilashvili, Davit; Gulbrandsen, Casper Fabian; Nguyen, Ha Thu; Havgar, Margrethe; Vedeler, Marianne & Hardeberg, Jon Yngve (2024). Toward Solving a Puzzle of Fragmented Archeological Textiles. Journal of Imaging Science and Technology. ISSN 1062-3701. 68(4), p. 1–16. doi: 10.2352/J.ImagingSci.Technol.2024.68.4.040407. Full text in Research Archive Show summary

  Archeological textiles can provide invaluable insight into the past. However, they are often highly fragmented, and a puzzle has to be solved to re-assemble the object and recover the original motifs. Unlike common jigsaw puzzles, archeological fragments are highly damaged, and no correct solution to the puzzle is known. Although automatic puzzle solving has fascinated computer scientists for a long time, this work is one of the first attempts to apply modern machine learning solutions to archeological textile re-assembly. First and foremost, it is important to know which fragments belong to the same object. Therefore, features are extracted from digital images of textile fragments using color statistics, classical texture descriptors, and deep learning methods. These features are used to conduct clustering and identify similar fragments. Four different case studies with increasing complexity are discussed in this article: from well-preserved textiles with available ground truth to an actual open problem of Oseberg archeological tapestry with unknown solution. This work reveals significant knowledge gaps in current machine learning, which helps us to outline a future avenue toward more specialized application-specific models.

• Gigilashvili, Davit; Lukesova, Hana & Hardeberg, Jon Yngve (2024). Criteria for matching fragmented archaeological textiles: a survey. Archaeological Textiles Review (ATR). ISSN 2245-7135. p. 64–75. Full text in Research Archive Show summary

  Archaeological textiles are a valuable source for understanding past cultures. However, textiles are vulnerable to decomposition and often found in a fragmentary state. Textile experts rely on their eyes and experience for manually matching similar fragments to reconstruct the original item. This requires much time and effort, as well as physical interaction with fragile artifacts. Computer-assisted virtual reconstruction has been recently proposed. Computer scientists suggested that the algorithms could mimic the criteria used by human experts. These criteria are rarely obvious or well-documented. To facilitate future interdisciplinary research efforts, a survey among archaeologists and conservators was conducted. This work explores what criteria are used by experts for matching fragments in the reconstruction process, which in turn may inspire the development of computational algorithms. Furthermore, the study evaluates the awareness among archaeologists and conservators of existing opportunities for computer-assisted reconstruction, their technological requirements and overall openness to such novel interdisciplinary solutions.

• Righetto, Leonardo; Khademizadeh, Mohammad; Giachetti, Andrea; Ponchio, Federico; Gigilashvili, Davit & Bettio, Fabio [Show all 7 contributors for this article] (2024). Efficient and User-Friendly Visualization of Neural Relightable Images for Cultural Heritage Applications. ACM Journal on Computing and Cultural Heritage. ISSN 1556-4673. 17(4), p. 1–24. doi: 10.1145/3690390. Full text in Research Archive Show summary

  We introduce an innovative multi-resolution framework for encoding and interactively visualizing large relightable images using a neural reflectance model derived from a state-of-the-art technique. The framework is seamlessly integrated into a scalable multi-platform framework that supports adaptive streaming and exploration of multi-layered relightable models in web settings. To enhance efficiency, we optimized the neural model, simplified decoding, and implemented a custom WebGL shader specific to the task, eliminating the need for deep-learning library integration in the code. Additionally, we introduce an efficient level-of-detail management system supporting fine-grained adaptive rendering through on-the-fly resampling in latent feature space. The resulting viewer facilitates interactive neural relighting of large images. Its modular design allows the incorporation of functionalities for cultural heritage analysis, such as loading and simultaneous visualization of multiple relightable layers with arbitra

• Gigilashvili, Davit; Lukesova, Hana; Gulbrandsen, Casper Fabian; Harijan, Akash & Hardeberg, Jon Yngve (2023). Computational techniques for virtual reconstruction of fragmented archaeological textiles. Heritage Science. 11(1). doi: 10.1186/s40494-023-01102-3. Full text in Research Archive Show summary

  Archaeological artifacts play important role in understanding the past developments of the humanity. However, the artifacts are often highly fragmented and degraded, with many details and parts missing due to centuries’ long degradation. Archaeologists and conservators attempt to reconstruct the original state of the objects either physically or virtually. This process includes characterizing and matching fragments’ features to identify which ones belong together. However, this process currently requires an extensive and tedious manual labor. Recent development in computational techniques gave rise to computer-assisted ways of virtual reconstruction, where the computer suggests solutions to the puzzle of scattered fragments and supplements or fully replaces manual labor. However, the capabilities of computational techniques remain limited in many aspects. This review summarizes the state-of-the-art computational techniques for puzzle and virtual reconstruction problems in cultural heritage applications, in general – with a particular interest in archaeological textiles. We overview existing computational methods, their applications and limitations. Afterward, based on the current knowledge gaps, we discuss where the field should go next.

• Gigilashvili, Davit; Nguyen, Ha Thu; Gulbrandsen, Casper Fabian; Havgar, Margrethe; Vedeler, Marianne & Hardeberg, Jon Yngve (2023). Texture-based clustering of archaeological textile images. Archiving Conference. ISSN 2161-8798. p. 139–142. doi: 10.2352/issn.2168-3204.2023.20.1.29. Full text in Research Archive Show summary

  Archaeological textiles are often highly fragmented, and solving a puzzle is needed to recover the original composition and respective motifs. The lack of ground truth and unknown number of the original artworks that the fragments come from complicate this process. We clustered the RGB images of the Viking Age Oseberg Tapestry based on their texture features. Classical texture descriptors as well as modern deep learning were used to construct a texture feature vector that was subsequently fed to the clustering algorithm. We anticipated that the clustering outcome would give indications to the number of original artworks. While the two clusters of different textures emerged, this finding needs to be taken with care due to a broad range of limitations and lessons learned.

• Lukesova, Hana & Kutzke, Hartmut (2023). Understanding of old treatments: Oseberg textile artworks studied from the conservation perspective. In Bridgland, Janet (Eds.), Working Towards a Sustainable Past. ICOM-CC 20th Triennial Conference Preprints, Valencia, 18-22 September 2023. International Council of Museums Committee for Conservation (ICOM-CC). ISSN 9788412754544. Full text in Research Archive Show summary

  Virtual Reconstruction, Interpretation and Preservation of the Textile Artifacts from the Oseberg Find, UiO

View all works in NVA

• Kutzke, Hartmut (2025). Ein archäologisches Puzzlespiel aus der Wikingerzeit: Die virtuelle Rekonstruktion der Wandteppiche des Osebergfundes. Full text in Research Archive
• Kutzke, Hartmut (2025). Virtual Reconstruction, Interpretation and Preservation of the Textile Artefacts from the Oseberg Find. Full text in Research Archive
• Gigilashvili, Davit (2025). Artifact Assembly and the Recent Advances in Oseberg Puzzle Solving. Full text in Research Archive
• Gigilashvili, Davit & Hardeberg, Jon Yngve (2025). When Pieces Don’t Fit: Why AI Still Struggles with Puzzle Solving. Full text in Research Archive Show summary

  Solving jigsaw puzzles is a fun pastime activity for humans. It is also a very intriguing problem from the computational perspective, and for the past seven decades, the computer scientists have been exploring different approaches for automatized puzzle reassembly. In addition to being an appealing challenge on its own, solving puzzles has an important role in archaeology and cultural heritage applications, where pieces need to be matched for virtual reconstruction and reassembly of the broken artifacts. We have recently conducted a literature review followed by set of experiments to discover that machine learning models struggle even on seemingly simple puzzle solving tasks. They often impose unrealistic simplifications and assumptions to successfully solve the puzzle. For instance, many models reassemble images that are split into regular grids of square pieces, while anything in real life hardly ever breaks, for instance, into 3-by-3 grid of perfect squares. This makes them of little use for real-life archaeological applications where the conditions of the puzzle are far from perfect: pieces come in highly irregular shapes, colors are faded, many areas are eroded and decomposed – thus, contours are not continuous, many pieces are missing altogether, and even with for the surviving pieces, it is not clear whether they belong to the same original object or how many different puzzles need to be solved in the first place. All of those can substantially complicate the computational solution of the puzzle, and existing state-of-the-art models systematically fail to overcome those limitations. What complicates the problem further is the lack of ground truth solutions for reliable evaluation of the results and the extremely limited specialized training datasets, because of which, the researchers should often resort to transfer learning. In this presentation, we will first present our benchmarking of state-of-the-art puzzle solvers on datasets of different difficulties; afterward, we propose an alternative approach – puzzle solving as a content-based image retrieval problem (CBIR), followed by CBIR benchmarking on different types of degradations and image irregularities. Finally, we will conclude with the overview of existing limitations, and discuss how machine learning should be best leveraged for complex puzzle solving. As an illustration, we will present how departure from transfer learning and using domain-specific training can improve archaeological puzzle solving.

• Gigilashvili, Davit (2025). Data science meets archaeology: a Viking puzzle. Full text in Research Archive
• Gigilashvili, Davit (2025). The Puzzling Problem of the Viking Textiles. Full text in Research Archive
• Khawaja, Muhammad Arsalan; Righetto, Leonardo; Gigilashvili, Davit; Storeide, Markus Sebastian Bakken; Trumpy, Giorgio & Hardeberg, Jon Yngve [Show all 7 contributors for this article] (2025). RTI for translucent beads. CAA2025: Digital Horizons: Embracing heritage in an evolving world. Full text in Research Archive
• Gigilashvili, Davit (2025). Artifact Assembly: What do the Vikings and jigsaw puzzles have in common? Full text in Research Archive
• Gigilashvili, Davit; George, Sony; Lukesova, Hana; Havgar, Margrethe; Vedeler, Marianne & Kutzke, Hartmut (2024). A Puzzling Problem of the Viking Age Tapestry. Full text in Research Archive
• Kutzke, Hartmut (2024). How to save a Viking treasure? Bio-inspired materials as a new generation of conservation materials for archaeological textiles – Keynote lecture. Full text in Research Archive
• Gigilashvili, Davit (2024). What do the Vikings and jigsaw puzzles have in common? Full text in Research Archive
• Kresovic, Milan; Gigilashvili, Davit; Hardeberg, Jon Yngve & Theoharis, Theoharis (2024). Virtual reconstruction of historical textiles using graph neural networks. Full text in Research Archive Show summary

  Archaeological artifacts are often found in a fragmentary state and archaeologists need to identify matching fragments and reassemble a puzzle to reconstruct the original and analyze its motifs. The process resembles jigsaw puzzle solving; puzzle solving is challenging for archaeological textiles, which are especially prone to fragmentation, degradation, fading and deformations [1]. A vivid example of such an open puzzle problem, is the Oseberg Tapestry from the Viking Age [6]. Previous works [2,3] used classical texture classifiers, as well as pre-trained neural networks to extract features from fragments and feed the feature vectors to clustering algorithms to identify similar fragments, but the performance was not deemed satisfactory (see [1] for full overview). In this work, we take an alternative approach and use Graph Neural Networks (GNN) for puzzle solving process. The eventual intended application is the case of the Oseberg Tapestry, which tells interesting stories from the Viking Age, but has many highly fragmented and missing pieces with irregular shapes [6]. However, due to its complexity and lack of complete reference, our study uses well-preserved, but stylistically similar textile. Using artificially fragmented textile allows us to evaluate the accuracy of puzzle solving, which is harder for the Oseberg Tapestry, due to lack of the ground truth. By artificially fragmenting the textile, we simulate an unrestricted pictorial puzzle [4], a type of puzzle where pieces are arbitrarily shaped, forming a complex planar adjacency graph that demands sophisticated reconstruction techniques. As mentioned in Harel and Ben-Shahar [4], there have been a few solutions for this type of puzzle, due to its complex nature. In recent years, one notable work comparable to our problem is a work on fragment assembly using graph neural network-based (GNN) reconstruction of papyrus fragments [5]. By virtually segmenting the textile imagery data (RGB and depth maps) into non-adjacent, different sizes of square patches, we construct a dataset needed for our GNN reconstruction model. The core of our reconstruction process involves extracting features from these tapestry patches using a pre-trained neural network. The extracted feature vectors form the nodes of fully connected graphs, encapsulating the relationships between tapestry segments. Using graph neural network’s pooling techniques, we can dynamically exchange information between nodes and update node embeddings containing visual and geometric data. Lastly, using the updated graph, we predict the edge connections information used to reconstruct the fragmented tapestry accurately. This methodology aims to predict the puzzle’s assembly, laying a foundation for future work in the virtual reconstruction of fragmented historical textiles.

• Fadillah, Riestiya Zain; Gigilashvili, Davit; Havgar, Margrethe; Vedeler, Marianne & Hardeberg, Jon Yngve (2024). Automatic Thread Counting for Archaeological Textiles. Full text in Research Archive Show summary

  Archaeological textile artifacts are often fragmented and human experts need to solve a puzzle to find matching pieces and reconstruct the original. This is a challenging task due to missing or degraded fragments, deformations and fading. Often it is not even known whether fragments come from the same object or not. A good example is the Oseberg Tapestry from the Viking Age [1]. Matching is usually done by experts who rely on technical analysis of the textile. One important criterion for identifying fragments from the same object is thread count (i.e. number of threads per centimeter) in the warp and weft directions. This is a tedious and time-consuming process. Computer science literature shows that current algorithms are not robust enough to solve the puzzle of archaeological textiles and proposes taking inspiration from the criteria that humans use [2-4]. To address this, we propose a framework for automatizing thread counting in textile images. We created a feature extractor, which captures thread count in vertical and horizontal directions. Thread count is usually measured by selecting several 1x1cm areas on the tapestries (we refer to this area as a patch), counting the threads and averaging the results. Archaeological textiles often have holes where threads are missing due to decomposition. The experts take this into account when they decide where to count. We modeled the thread count extraction as a pulse-counting process, inspired by how we intuitively count threads. For counting vertical threads, we trace the patch in the horizontal direction and count the thread by finding the number of occurrences of two consecutive vertical edges. Similarly for the horizontal threads we trace the patch in vertical directions. Our proposed thread flow begins with applying Otsu’s binary thresholding to the patch. Then we take all the pixel values for all columns in the same row to model the horizontal tracing of the patch. The pulse front is obtained by calculating the absolute approximate derivative of the trace. The number of pulses, which represent the number 𝑋(2: 𝑛) − 𝑋(1: 𝑛−1) of thread, is obtained by the number of non-zero elements of the absolute approximate derivative divided by two. This operation is repeated all the way from the top to bottom patch and the majority thread count is returned as the thread density of the patch. We compared our algorithm with the ground truth counted by experts. The preliminary results show that our method overestimates thread count. The final results will be presented at the conference. However, thread count features extracted with our method, provided promising results when used for clustering. Twelve out of 27 fragments clustered using thread count alone, were aligned with archaeologists’ hypotheses on which fragments belong together. The Oseberg Tapestry is a challenging case-study also because its thread systems consist of double threads, which introduces an additional complexity to the machine. The result for simpler scenarios will be also presented. Future work should explore more sophisticated methods for thresholding, such as Sobel to expose edges, and STRESS (Spatio-Temporal Retinex-inspired Envelope with Stochastic Sampling) enhancement preprocessing, to make them more separable.

• Gigilashvili, Davit (2024). What do the Vikings and jigsaw puzzles have in common? Full text in Research Archive
• Gigilashvili, Davit (2024). Virtual Reconstruction of Highly Fragmented Archaeological Textiles. Full text in Research Archive
• Gigilashvili, Davit (2024). Software for Virtual Reconstruction of Cultural Heritage Artifacts. Full text in Research Archive
• Gigilashvili, Davit (2024). Toward Solving a Puzzle of Fragmented Archeological Textiles. Full text in Research Archive
• Gigilashvili, Davit (2024). Ny viten om vev fra vikingtiden. [Internet]. Gemini. Full text in Research Archive
• Gigilashvili, Davit (2024). Avslører eldgammel hemmelighet. [Journal]. Oppland Arbeiderblad. Full text in Research Archive
• Gigilashvili, Davit (2024). På Gjøvik puslar dei saman stoffbitar funne i vikinggrav saman med Osebergskipet. [TV]. NRK. Full text in Research Archive
• Gigilashvili, Davit (2024). Pusler sammen 1200 år gamle stoffbiter fra Osebergskipet. [TV]. NRK. Full text in Research Archive
• Gigilashvili, Davit (2024). Piecing together Viking tapestries from the Oseberg Ship. [Internet]. Norwegian SciTech News. Full text in Research Archive
• Gigilashvili, Davit (2024). Ni menn som kan ha blitt ofret – moderne teknologi skal avsløre mer fra teppene i Oseberg-graven. [Internet]. forskning.no. Full text in Research Archive
• Gigilashvili, Davit (2024). Skjulte fortællinger fra Osebergskibet kommer frem i lyset. [Internet]. Historienet.dk. Full text in Research Archive
• Gigilashvili, Davit (2024). Software for Virtual Reconstruction in Cultural Heritage Applications. Full text in Research Archive
• Gigilashvili, Davit (2024). Skjulte fortellinger fra Osebergskipet kommer frem i lyset. [Internet]. Historie. Full text in Research Archive
• Wakefield, Jennifer; Jones, Chas; Lukesova, Hana; Harding, Stephen E & Kutzke, Hartmut (2024). Archaeological woollen textiles in a fragile state: the potential of polysaccharide based consolidants. Full text in Research Archive
• Schadt, Melina; Pöschke, C; Wiertz, Fabian; Dittrich, Barbara; Kutzke, Hartmut & Herrmann, Andreas (2024). Bio-based supercharged protein glue for cultural heritages conservation. Full text in Research Archive
• Flo, Line; Kutzke, Hartmut; Meulen, Douwtje Lieuwkje van der & Nabais, Paula (2024). The Poor Persons Purple. Investigations into a lichen Orcein dye from Norwegian domestic lichen. Full text in Research Archive
• Wakefield, Jennifer; Flo, Line; Jones, Chas; Harding, Stephen E & Kutzke, Hartmut (2024). The Potential of Orbi-SIMS for Investigating Dyes in Archaeological Textiles. Full text in Research Archive
• Wakefield, Jennifer; Flo, Line; Jones, Chas; Lukesova, Hana; Harding, Stephen E & Kutzke, Hartmut (2024). Investigating Polysaccharide-based Consolidates for fragile archaeological woollen textiles. Full text in Research Archive
• Gigilashvili, Davit (2024). Reconstructing the Past: Puzzle of Viking Textiles. Full text in Research Archive
• Gigilashvili, Davit (2024). Artifact Assembly for Puzzle Solving. Full text in Research Archive
• Kutzke, Hartmut (2023). Learning from nature: developing bio-inspired and biomimetic materials as a new generation of materials for preservation of cultural heritage – keynote lecture. Full text in Research Archive
• Gigilashvili, Davit (2023). Virtual Reconstruction of Highly Fragmented Archaeological Textiles – A Puzzle Problem. Full text in Research Archive
• Lukesova, Hana & Kutzke, Hartmut (2023). Understanding of old treatments: Oseberg textile artworks studied from the conservation perspective. Full text in Research Archive Show summary

  Virtual Reconstruction, Interpretation and Preservation of the Textile Artifacts from the Oseberg Find, UiO

• Gigilashvili, Davit (2023). Appearance Beyond Colour: Gloss, Translucency, and More. Full text in Research Archive
• Dittrich, Barbara; Thomas, Tina-Marie; Lukesova, Hana; Kutzke, Hartmut & Herrmann, Andreas (2022). Bioinspired glues for the conservation of archaeological textile artefacts. Full text in Research Archive
• Kutzke, Hartmut; Lukesova, Hana & Vedeler, Marianne (2022). TexRec – a project on studying and virtually reconstructing the Viking-age Oseberg tapestries. Full text in Research Archive
• Kool, Anneleen (2021). Plant dyes and historical botany. Full text in Research Archive
• Kutzke, Hartmut (2021). The TexRec project. Full text in Research Archive

View all works in NVA