Center for Cultural Heritage Technology

The application of Machine Learning (ML) to Cultural Heritage (CH) has evolved since basic statistical approaches such as Linear Regression to complex Deep Learning models. The question remains how much of this actively improves on the underlying algorithm versus using it within a ‘black box’ setting. We survey across ML and CH literature to identify the theoretical changes which contribute to the algorithm and in turn them suitable for CH applications. Alternatively, and most commonly, when there are no changes, we review the CH applications, features and pre/post-processing which make the algorithm suitable for its use. We analyse the dominant divides within ML, Supervised, Semi-supervised and Unsupervised, and reflect on a variety of algorithms that have been extensively used. From such an analysis, we give a critical look at the use of ML in CH and consider why CH has only limited adoption of ML.

Given the multiplicity of nanoparticles (NPs), there is a requirement to develop screening strategies to evaluate their toxicity. Within the EU-funded FP7 NanoTEST project, a panel of medically relevant NPs has been used to develop alternative testing strategies of NPs used in medical diagnostics. As conventional toxicity tests cannot necessarily be directly applied to NPs in the same manner as for soluble chemicals and drugs, we determined the extent of interference of NPs with each assay process and components. In this study, we fully characterized the panel of NP suspensions used in this project (poly(lactic-co-glycolic acid)-polyethylene oxide [PLGA-PEO], TiO2, SiO2, and uncoated and oleic-acid coated Fe3O4) and showed that many NP characteristics (composition, size, coatings, and agglomeration) interfere with a range of in vitro cytotoxicity assays (WST-1, MTT, lactate dehydrogenase, neutral red, propidium iodide, (3)H-thymidine incorporation, and cell counting), pro-inflammatory response evaluation (ELISA for GM-CSF, IL-6, and IL-8), and oxidative stress detection (monoBromoBimane, dichlorofluorescein, and NO assays). Interferences were assay specific as well as NP specific. We propose how to integrate and avoid interference with testing systems as a first step of a screening strategy for biomedical NPs.

In the last decades, the interest in the development of protective coatings for movable and immovable Cultural Heritage (CH) assets has decidedly increased. This has been mainly prompted by the raising consciousness on preservation requirements for cultural artefacts and monuments, which has consequently determined the development of new protective products. From acrylic resins used at the end of the last century to the up-to-date biomaterials and nanoparticles employed nowadays, the research has made a giant step forward. This article reviews the progresses, the technical challenges, and the most recent advances in protective coatings for archaeological metal, glass, and stone artefacts. It aims at offering a comprehensive and critical overview of the progressions in conservation science and displaying how research has optimized polymers in order to solve deterioration problems. Attention is given to recently developed materials, hybrid coatings, and corrosion inhibitors. This work seeks to provide a reference point for future research and to offer a wide-ranging introduction on the newly available material technologies to restorers and conservators.

Wetlands provide vital ecological and socioeconomic services but face escalating pressures worldwide. This study undertakes an integrated spatiotemporal assessment of the multifaceted vulnerabilities shaping Khinjhir Lake, an ecologically significant wetland ecosystem in Pakistan, using advanced geospatial and machine learning techniques. Multi-temporal optical remote sensing data from 2000 to 2020 was analyzed through spectral water indices, land cover classification, change detection and risk mapping to examine moisture variability, land cover modifications, area changes and proximity-based threats over two decades. The random forest algorithm attained the highest accuracy (89.5%) for land cover classification based on rigorous k-fold cross-validation, with a training accuracy of 91.2% and a testing accuracy of 87.3%. This demonstrates the model’s effectiveness and robustness for wetland vulnerability modeling in the study area, showing 11% shrinkage in open water bodies since 2000. Inventory risk zoning revealed 30% of present-day wetland areas under moderate to high vulnerability. The cellular automata–Markov (CA–Markov) model predicted continued long-term declines driven by swelling anthropogenic pressures like the 29 million population growth surrounding Khinjhir Lake. The research demonstrates the effectiveness of integrating satellite data analytics, machine learning algorithms and spatial modeling to generate actionable insights into wetland vulnerability to guide conservation planning. The findings provide a robust baseline to inform policies aimed at ensuring the health and sustainable management and conservation of Khinjhir Lake wetlands in the face of escalating human and climatic pressures that threaten the ecological health and functioning of these vital ecosystems.

This study undertook an assessment of 24 physiochemical parameters at over 1094 sites to compute the water quality index (WQI) across the upper and central Punjab regions of Pakistan. Prior to the WQI calculation, an analytical hierarchy process (AHP) was employed to assign specific weights to each water quality parameter. The categorization of WQI into distinct classes was achieved by constructing a pairwise matrix based on their relative importance utilizing Saaty’s scale. Additionally, the groundwater quality status for irrigation and drinking purposes across various zones in the study area was delineated through the integration of WQI and geostatistical methodologies. The findings revealed discernible heavy metal issues in the Lahore division, with emerging microbiological contamination across the entire study region, potentially attributed to untreated industrial effluent discharge and inadequately managed sewerage systems. The computed indices for the Lahore, Sargodha, and Rawalpindi divisions fell within the marginal to unfit categories, indicating water quality concerns. In contrast, the indices for other divisions were in the medium class, suggesting suitability for drinking purposes. Scenario analysis for developing mitigation strategies indicated that primary treatment before wastewater disposal could rehabilitate 9% of the study area, followed by secondary (35%) and tertiary (41%) treatments. Microbiological contamination (27%) emerged as the predominant challenge for water supply agencies. Given the current trajectory of water quality deterioration, access to potable water is poised to become a significant public concern. Consequently, government agencies are urged to implement appropriate measures to enhance overall groundwater quality for sustainable development.

Abstract Glass has been used in widespread applications within several sectors since ancient times and it has been systematically studied under different perspectives. However, its thermodynamic properties and the variety of its compositions, several aspects related to its durability and its alteration mechanisms remain still open to debate. This literature review presents an overview of the most relevant studies on glass corrosion and the interaction between glass and the environment. The review aims to achieve two objectives. On one hand, it aims to highlight how far research on glass corrosion has come by studying model systems created in the laboratory to simulate different alteration conditions and glass compositions. On the other, it seeks to point out what are the critical aspects that still need to be investigated and how the study of ancient, altered glass can add to the results obtained in laboratory models. The review intends also to demonstrate how advanced analytical techniques commonly used to study modern and technical glass can be applied to investigate corrosion marks on ancient samples.

Abstract Cultural heritage faces recurring degradation processes and natural aging phenomena, demanding the envisioning of innovative preservation solutions inspired by cutting‐edge scientific research. Over extended time frames, current preservation strategies often prove inadequate in preserving the different constituent materials of cultural assets, which are thus threatened by their inherent fragility and by the complex interactions with the surrounding environment. The distinctive properties of graphene and graphene‐related materials (GRMs) now offer unexplored opportunities in the field of cultural heritage, addressing various forms of deterioration phenomena. This work critically analyzes early‐stage literature on the use of graphene and GRMs. Strengths, weaknesses, and limitations in anti‐corrosion, anti‐fading, and consolidation properties of graphene and GRMs are thoroughly investigated, along with their possible applications in smart sensors to monitor the state of health of endangered artifacts. The aim is to elucidate how specific characteristics of graphene and GRMs can be applied to the conservation, diagnostics, and monitoring of artistic and archaeological assets. Future perspectives in the design of stable, long‐lasting, and compatible graphene‐based solutions for cultural heritage protection are highlighted, providing a detailed discussion on potentials and pitfalls.

We propose a novel information gain metric that combines hand-crafted and data-driven metrics to address the next best view problem for autonomous 3-D mapping of unknown indoor environments. For the hand-crafted metric, we propose an entropy-based information gain that accounts for the previous view points to avoid the camera to revisit the same location and to promote the motion toward unexplored or occluded areas. However, for the learnt metric, we adopt a convolutional neural network (CNN) architecture and formulate the problem as a classification problem. The CNN takes the current depth image as input and outputs the motion direction that suggests the largest unexplored surface. We train and test the CNN using a new synthetic dataset based on the SUNCG dataset. The learnt motion direction is then combined with the proposed hand-crafted metric to help handle situations where using only the hand-crafted metric tends to face ambiguities. We finally evaluate the autonomous paths over several real and synthetic indoor scenes including complex industrial and domestic settings and prove that our combined metric is able to further improve the exploration coverage compared to using only the proposed hand-crafted metric.

Machine Learning-based workflows are being progressively used for the automatic detection of archaeological objects (intended as below-surface sites) in remote sensing data. Despite promising results in the detection phase, there is still a lack of a standard set of measures to evaluate the performance of object detection methods, since buried archaeological sites often have distinctive shapes that set them aside from other types of objects included in mainstream remote sensing datasets (e.g., Dataset of Object deTection in Aerial images, DOTA). Additionally, archaeological research relies heavily on geospatial information when validating the output of an object detection procedure, a type of information that is not normally considered in regular machine learning validation pipelines. This paper tackles these shortcomings by introducing two novel automatic evaluation measures, namely ‘centroid-based’ and ‘pixel-based’, designed to encode the salient aspects of the archaeologists’ thinking process. To test their usability, an experiment with different object detection deep neural networks was conducted on a LiDAR dataset. The experimental results show that these two automatic measures closely resemble the semi-automatic one currently used by archaeologists and therefore can be adopted as fully automatic evaluation measures in archaeological remote sensing detection. Adoption will facilitate cross-study comparisons and close collaboration between machine learning and archaeological researchers, which in turn will encourage the development of novel human-centred archaeological object detection tools.

“Tells” are archaeological mounds formed by deposition of large amounts of anthropogenic material and sediments over thousands of years and are the most important and prominent features in Near and Middle Eastern archaeological landscapes. In the last decade, archaeologists have exploited free-access global digital elevation model (DEM) datasets at medium resolution (i.e., up to 30 m) to map tells on a supra-regional scale and pinpoint tentative tell sites. Instead, the potential of satellite DEMs at higher resolution for this task was yet to be demonstrated. To this purpose, the 3 m resolution imaging capability allowed by the Italian Space Agency’s COSMO-SkyMed Synthetic Aperture Radar (SAR) constellation in StripMap HIMAGE mode was used in this study to generate DEM products of enhanced resolution to undertake, for the first time, a systematic mapping of tells and archaeological deposits. The demonstration is run at regional scale in the Governorate of Wasit in central Iraq, where the literature suggested a high density of sites, despite knowledge gaps about their location and spatial distribution. Accuracy assessment of the COSMO-SkyMed DEM is provided with respect to the most commonly used SRTM and ALOS World 3D DEMs. Owing to the 10 m posting and the consequent enhanced observation capability, the COSMO-SkyMed DEM proves capable to detect both well preserved and levelled or disturbed tells, standing out for more than 4 m from the surrounding landscape. Through the integration with CORONA KH-4B tiles, 1950s Soviet maps and recent Sentinel-2 multispectral images, the expert-led visual identification and manual mapping in the GIS environment led to localization of tens of sites that were not previously mapped, alongside the computation of a figure as up-to-date as February 2019 of the survived tells, with those affected by looting. Finally, this evidence is used to recognize hot-spot areas of potential concern for the conservation of tells. To this purpose, we upgraded the spatial resolution of the observations up to 1 m by using the Enhanced Spotlight mode to collect a bespoke time series. The change detection tests undertaken on selected clusters of disturbed tells prove how a dedicated monitoring activity may allow a regular observation of the impacts due to anthropogenic disturbance (e.g., road and canal constructions or ploughing).

The increasing need of restoring high-resolution hyper-spectral (HS) images is determining a growing reliance on computer vision-based processing to enhance the clarity of the image content. HS images can, in fact, suffer from degradation effects or artefacts caused by instrument limitations. This article focuses on a procedure aimed at reducing the degradation effects, frequency-dependent blur and noise, in Terahertz time-domain spectroscopy (THz-TDS) images in reflection geometry. It describes the application of a joint deblurring and denoising approach that had been previously proved to be effective for the restoration of THz-TDS images in transmission geometry, but that had never been tested in reflection modality. This mode is often the only one that can be effectively used in most cases, for example, when analyzing objects that are either opaque in the THz range, or that cannot be displaced from their location (e.g., museums), such as those of cultural interest. Compared to transmission mode, reflection geometry introduces, however, further distortion to THz data, usually neglected in existing literature. In this work, we successfully implement image deblurring and denoising of both uniform-shape samples (a contemporary 1 Euro cent coin and an inlaid pendant) and samples with the uneven reliefs and corrosion products on the surface, which make the analysis of the object particularly complex (an ancient Roman silver coin). The study demonstrates the ability of image processing to restore data in the 0.25–6 THz range, spanning over more than four octaves, and providing the foundation for future analytical approaches of cultural heritage using the far-infrared spectrum still not sufficiently investigated in the literature.

Flood susceptibility prediction is complex due to the multifaceted interactions among hydrological, meteorological, and urbanisation factors, further exacerbated by climate change. This study addresses these complexities by investigating flood susceptibility in rapidly urbanising regions prone to extreme weather events, focusing on Gdańsk, Poland. Three popular ML techniques, Support Vector Machine (SVM), Random Forest (RF), and Artificial Neural Networks (ANN), were evaluated for handling complex, nonlinear data using a dataset of 265 urban flood episodes. An ensemble filter feature selection (EFFS) approach was introduced to overcome the single-method feature selection limitations, optimising the selection of factors contributing to flood susceptibility. Additionally, the study incorporates explainable artificial intelligence (XAI), namely, the Shapley Additive exPlanations (SHAP) model, to enhance the transparency and interpretability of the modelling results. The models’ performance was evaluated using various statistical measures on a testing dataset. The ANN model demonstrated a superior performance, outperforming the RF and the SVM. SHAP analysis identified rainwater collectors, land surface temperature (LST), digital elevation model (DEM), soil, river buffers, and normalized difference vegetation index (NDVI) as contributors to flood susceptibility, making them more understandable and actionable for stakeholders. The findings highlight the need for tailored flood management strategies, offering a novel approach to urban flood forecasting that emphasises predictive power and model explainability.

The ability of shape-controlled octahedral Pt nanoparticles to act as nanozyme mimicking glucose oxidase enzyme is reported. Extended {111} particle surface facets coupled with a size comparable to natural enzymes and easy-to-remove citrate coating give high affinity for glucose, comparable to the enzyme as proven by the steady-state kinetics of glucose electrooxidation. The easy and thorough removal of the citrate coating, demonstrated by X-ray photoelectron spectroscopy analysis, allows a highly stable deposition of the nanozymes on the electrode. The glucose electrochemical detection (at -0.2 V vs SCE) shows a linear response between 0.36 and 17 mM with a limit of detection of 110 μM. A good reproducibility has been achieved, with an average relative standard deviation (RSD) value of 9.1% (n = 3). Similarly, a low intra-sensor variability has been observed, with a RSD of 6.6% (n = 3). Moreover, the sensor shows a long-term stability with reproducible performances for at least 2 months (RSD: 7.8%). Tests in saliva samples show the applicability of Pt nanozymes to commercial systems for non-invasive monitoring of hyperglycemia in saliva, with recoveries ranging from 92 to 98%.

The historical knowledge inherited from house paint documents and the experimental research on synthetic pigments show that production methods have an important role in the performance of paint. In this regard, this work investigates the links existing between the optical emission, crystal defects and photocatalytic activity of zinc white pigment from different contemporary factories, with the aim of elucidating the effects of these characteristics onto the tendency of the pigment to induce paint failures. The analysed samples display highly similar crystallite structure, domain size, and specific surface area, whilst white pigments differ from pure ZnO in regards to the presence of zinc carbonate hydrate that is found as a foreign compound. In contrast, the photoluminescence measurements categorize the analysed samples into two groups, which display different trap-assisted emissions ascribed to point crystal defects introduced during the synthesis process, and associated to Zn or O displacement. The photocatalytic degradation tests infer that the emerged defective structure and specific surface area of ZnO-based samples influence their tendency to oxidize organic molecules under light irradiation. In particular, the results indicate that the zinc interstitial defects may be able to promote the photogenerated electron-hole couples separation with a consequent increase of the overall ZnO photocatalytic activity, negatively affecting the binding medium stability. This groundwork paves the way for further studies on the link between the photoluminescence emission of the zinc white pigment and its tendency to decompose organic components contained in the binding medium.

The growth of pyramidal platinum nanocrystals is studied by a combination of synthesis/characterization experiments and density functional theory calculations. It is shown that the growth of pyramidal shapes is due to a peculiar type of symmetry breaking, which is caused by the adsorption of hydrogen on the growing nanocrystals. Specifically, the growth of pyramidal shapes is attributed to the size-dependent adsorption energies of hydrogen atoms on {100} facets, whose growth is hindered only if they are sufficiently large. The crucial role of hydrogen adsorption is further confirmed by the absence of pyramidal nanocrystals in experiments where the reduction process does not involve hydrogen.

Terahertz (THz) pulse/time-domain imaging attracted increased interest in recent years mostly due to its ability to extract dielectric properties of sample materials (i.e., absorption coefficient and the refraction index) from the amplitude and phase of each spectral component of the THz pulse. The resulting data from a THz time-domain system represents a 3-dimensional (3D) hyperspectral cube which contains several 2D images corresponding to different frequencies or bands. Due to a frequency-dependent non-zero THz beam waist, these 2D images are corrupted by blurring artifacts: a THz beam waist is wider on lower frequencies leading to more blurry corresponding 2D images. At higher frequencies, the beam waist is smaller resulting in sharper, but noisier images due to the decrease in the THz signal amplitude. The main focus of this work is the joint reduction of blur and noise from THz time-domain images. We propose two instances of a fast joint deblurring and denoising approach which is able to deal with THz time-domain images corrupted by different noise types and frequency-dependent blur. The experiments performed on synthetic and real THz time-domain images show that the proposed approach outperforms conventional 2D deblurring approaches and methods tailored to remote sensing hyperspectral images. To the best of our knowledge, this is the first time that a joint deblurring and denoising approach tailored to THz time-domain images is proposed taking into consideration band-dependent blur and different noise types.

Historical paper documents are susceptible to complex degradation processes, including biodeterioration, which can progressively compromise their aesthetic and structural integrity. This study analyses seventeenth century handwritten historical letters stored at the Correr Museum Library in Venice, Italy, exhibiting pronounced signs of biodeterioration. The techniques used encompassed traditional colony isolation on agar plates and proteomics analyses, employing nanoscale liquid chromatography coupled with high-resolution mass spectrometry (nano-LC-MS). Fluorescence microscopy was used for the first time in the historical paper biodeterioration context to supplement the conventional stereoscopic, optical, and scanning electron microscopic imaging techniques. This method enables the visualisation of microorganisms beyond and beneath the paper's surface through their natural intrinsic autofluorescence in a non-invasive and non-destructive way. The results demonstrate a diverse, complex, and abundant microbiota composed of coexisting fungal and bacterial species (Ascomycota, Mucoromycota, Basidiomycota, Proteobacteria, and Actinobacteria), along with mite carcasses, insects, parasites, and possibly protists. Furthermore, this study reveals certain species that were not previously documented in the biodeterioration of historical paper, including human pathogens, such as Histoplasma capsulatum, Brucella, Candida albicans, and species of Aspergillus (A. flavus, A. fumigatus, A. oryzae, A. terreus, A. niger) known to cause infections or produce mycotoxins, posing substantial risk to both artefacts and humans.

Ancient glass objects typically show distinctive effects of deterioration as a result of environmentally induced physicochemical transformations of their surface over time. Iridescence is one of the distinctive signatures of aging that is most commonly found on excavated glass. In this work, we present an ancient glass fragment that exhibits structural color through surface weathering resulting in iridescent patinas caused by silica reprecipitation in nanoscale lamellae. This archaeological artifact reveals an unusual hierarchically assembled photonic crystal with extremely ordered nanoscale domains, high spectral selectivity, and reflectivity (~90%), that collectively behaves like a gold mirror. Optical characterization paired with nanoscale elemental analysis further underscores the high quality of this structure providing a window into this sophisticated natural photonic crystal assembled by time.

Abstract. This paper introduces a novel methodology developed for creating 3D models of archaeological artifacts that reduces the time and effort required by operators. The approach uses a simple vision system mounted on a robotic arm that follows a predetermined path around the object to be reconstructed. The robotic system captures different viewing angles of the object and assigns 3D coordinates corresponding to the robot's pose, allowing it to adjust the trajectory to accommodate objects of various shapes and sizes. The angular displacement between consecutive acquisitions can also be fine-tuned based on the desired final resolution. This flexible approach is suitable for different object sizes, textures, and levels of detail, making it ideal for both large volumes with low detail and small volumes with high detail. The recorded images and assigned coordinates are fed into a constrained implementation of the structure-from-motion (SfM) algorithm, which uses the scale-invariant features transform (SIFT) method to detect key points in each image. By utilising a priori knowledge of the coordinates and SIFT algorithm, low processing time can be ensured while maintaining high accuracy in the final reconstruction.The use of a robotic system to acquire images at a pre-defined pace ensures high repeatability and consistency across different 3D reconstructions, eliminating operator errors in the workflow. This approach not only allows for comparisons between similar objects but also provides the ability to track structural changes of the same object over time.Overall, the proposed methodology provides a significant improvement over current photogrammetry techniques by reducing the time and effort required to create 3D models while maintaining a high level of accuracy and repeatability.

This work demonstrates terahertz time-domain spectroscopy (THz-TDS) in reflection configuration on a class of inorganic and mineral pigments. The technique is validated for pictorial materials against the limitations imposed by the back-reflection of the THz signal, such as weak signal intensity, multiple signal losses and distortion, as well as the current scarce databases. This work provides a detailed description of the experimental procedure and method used for the determination of material absorption coefficient of a group of 10 pigments known to be used in ancient frescoes, that are, Cu-based (azurite, malachite, and Egyptian blue), Pb-based (minium and massicot), Fe-based (iron oxide yellow, dark ochre, hematite, and Pompeii red) pigments and mercury sulfide (cinnabar), and classified the vibrational modes of the molecular oxides and sulfides for material identification. The results of this work showed that the mild signal in reflection configuration does not limit the application of THz-TDS on inorganic and mineral pigments as long as (i) the THz signal is normalized with a highly reflective reference sample, (ii) the secondary reflected signals from inner interfaces are removed with a filtering procedure, and (iii) the limitations at high frequencies imposed by the dynamic range of the instrument are considered. Under these assumptions, we were able to differentiate molecular phases of the same metal and identify azurite, Egyptian blue, minium, and cinnabar, isolating the molecular vibrations up to 125 cm −1 . The established approach demonstrated to be reliable, and it can be extended for the study of other materials, well beyond the reach of the heritage domain.