Next Article in Journal
Architecture and Armour in Heritage Discourse: Form, Function, and Symbolism
Previous Article in Journal
The Characterization and Delineation of the Material Cultural Heritage Sector in Europe
 
 
Search for Articles:
Title / Keyword
Author / Affiliation / Email
Journal
Article Type
 
 
Advanced Search
 
Section
Special Issue
Volume
Issue
Number
Page
 
Journals
Heritage
Volume 8
Issue 9
10.3390/heritage8090381
heritage-logo
Submit to this Journal Review for this Journal Propose a Special Issue
► Article Menu

Article Menu

share Share announcement Help format_quote Cite question_answer Discuss in SciProfiles
first_page
settings
Order Article Reprints
Font Type:
Arial Georgia Verdana
Font Size:
Aa Aa Aa
Line Spacing:
Column Width:
Background:
Review

Development Trend in Non-Destructive Techniques for Cultural Heritage: From Material Characterization to AI-Driven Diagnosis

by
Mingrui Zhang
1,2,
Suchi Liu
1,
Haojian Shao
1,
Zonghuan Ba
1,
Jie Liu
1,2,*,
Mǎdǎlina Georgiana Albu Kaya
3,
Keyong Tang
1,2,* and
Guohe Han
2,*
1
School of Materials Science and Engineering, Zhengzhou University, Zhengzhou 450001, China
2
Archaeology Innovation Center, Zhengzhou University, Zhengzhou 450001, China
3
Collagen Department, INCDTP-Leather and Footwear Research Institute, 93 Ion Minulescu, 031215 Bucharest, Romania
*
Authors to whom correspondence should be addressed.
Heritage 2025, 8(9), 381; https://doi.org/10.3390/heritage8090381
Submission received: 22 July 2025 / Revised: 22 August 2025 / Accepted: 8 September 2025 / Published: 16 September 2025
Browse Figures
Versions Notes

Abstract

Cultural heritage (CH) relics are irreplaceable records of human civilization, encompassing diverse historical, technological, and artistic achievements. Extracting their structural and compositional information without affecting their physical integrity is a critical challenge. This review summarizes recent advances in non-destructive techniques (NDTs) for CH analysis and emphasizes the balance between the depth of analysis and conservation ethics. Techniques are broadly categorized into spectrum-based, X-ray-based, and digital-based methods. Spectroscopic techniques such as Fourier transform infrared (FTIR), Raman, and nuclear magnetic resonance (NMR) spectroscopy provide molecular-level insights into organic and inorganic components, often requiring minimal or no sampling. X-ray-based techniques, including conventional and spatially resolved XRD/XRF and total reflection XRF (TRXRF), provide powerful means for crystal and elemental analysis, including in situ pigment identification and trace material analysis. Digital-based methods include high-resolution imaging, three-dimensional modeling, data fusion, and AI-driven diagnosis to achieve the non-invasive visualization, monitoring, and virtual restoration of CH assets. This review highlights a methodology shift from traditional molecular-level detection to data-centric and AI-assisted diagnosis, reflecting the paradigm shift in heritage science.

1. Introduction

Cultural heritage (CH) is a tangible record of human civilization, reflecting the historical lifestyle, technology, and artistic achievements of different periods and regions [1]. Cultural relics, such as ancient buildings [2], vessels [3], manuscripts [4], and costumes [5], could provide insight into environmental changes and regional customs. For example, the murals in Egyptian pyramids reveal the social structure and religious beliefs of ancient Egypt [6], while the Buddhist murals in the Mogao Caves (Dunhuang, China) provide valuable information about the culture and society of the Tang Dynasty [7]. In addition to their historical and esthetic values, CH strengthens collective identity and fosters emotional ties with their ancestral roots. Therefore, the protection of these cultural heritages has become the global responsibility of researchers, protectors, and institutions.
However, CH preservation faces a double challenge: how does one accurately extract the composition and structure information of cultural relics without affecting their physical integrity? In the real world, this tension is often obvious. Many objects are extremely fragile, unique, or irreplaceable, severely restricting or completely prohibiting physical sampling. For example, polychrome terracotta figures were usually coated with pigment layers of several microns thick, so sampling may lead to irreversible surface loss. Similarly, historical paper artifacts under long-term humidity fluctuations are usually fragile and powdery and will disintegrate even under the minimum mechanical stress. In museums and archeological sites, many high-value items are stored in sealed environments or exhibition boxes, where contact is strictly prohibited. These cases reflect a persistent and unresolved tension between the demands of scientific analysis and the inviolable principles of CH conservation.
Physical sampling is no longer permitted at major heritage sites like the Lascaux Caves, Nefertari’s Tomb, Pompeii, Göreme churches, Persepolis, and Masada, as documented in institutional archives. However, historical samples preserved in archives such as those held by ICCROM can be re-examined using modern non-destructive analytical techniques. This approach provides valuable new insights while fully safeguarding these irreplaceable heritage materials. In this context, non-destructive techniques (NDTs) have become indispensable tools for CH diagnosis and conservation. In view of the irreplaceable and fragile nature of CH, it is essential to minimize physical intervention in the analysis process. For immovable or delicate objects, such as building components, large sculptures, or murals, portable instruments such as those used in X-ray fluorescence (XRF), Raman spectroscopy, and handheld X-ray diffraction (XRD) are becoming more and more popular because of their mobility and ease of operation [8,9,10]. These techniques allow for the in situ, non-contact characterization of cultural materials and support conservation decisions without changing artifacts.
Despite its increasingly wide application, the definition of “non-destructive” has subtle differences in CH research. Some scholars insist on strict interpretation, requiring zero sampling without changing the object surface [11], while others adopt a more flexible view, allowing micro-sampling without affecting the overall condition of the object [12]. Although the former conforms to the concept of protection, the diversity and complexity of cultural materials often require minimally invasive methods. Fortunately, the improvement in analysis enables many techniques to be applied to trace samples. For example, micro-ATR-FTIR can detect surface compositions using particles in the range of milligrams or sub-milligrams, while high-resolution XRD can characterize the structure from microgram powder. This review adopts a pragmatic position, focusing on non-destructive or minimally invasive methods without compromising the integrity or value of the materials studied.
In addition to the technical level, NDTs also play a broader cultural and ethical role. CH is not only the object of a scientific research but also the carrier of collective memory and human identity. Enhancing our ability to study these materials without causing harm reflects our commitment to cultural respect and sustainability. In addition, non-destructive diagnosis promotes interdisciplinary collaboration across science, conservation, and humanities and contributes to global cultural exchanges and mutual understanding. In this sense, the development and application of NDTs not only represent technological progress, but they are also a cultural need. To align technological advances with cultural responsibilities, it is essential to ground NDT applications in established ethical and procedural frameworks.
The adoption of non-destructive or minimally invasive approaches in CH research is underpinned by widely recognized ethical and procedural frameworks. International bodies such as the International Council of Museums (ICOM), the American Institute for Conservation (AIC), UNESCO, and ICCROM stipulate in their Codes of Ethics that sampling should follow the principles of minimal intervention, prior informed consent from the owner or custodian, and full documentation of procedures and results [13]. The Canadian Conservation Institute (CCI) and the Canadian Association for Conservation of Cultural Property (CAC) similarly emphasize shared responsibility, transparency, and the minimization of physical impact in conservation practice. At the procedural level, the European Standard EN 16085:2012 (“Conservation of cultural property—Methodology for sampling from materials of cultural property—General rules”) provides explicit requirements for justifying, authorizing, and documenting any sampling, including criteria for sample size, representativeness, and chain-of-custody management [14]. Complementary to this, EN 16853:2017 defines the decision-making, planning, and implementation process for conservation interventions [15], while EN 16095:2012 offers standardized protocols for pre- and post-sampling condition reporting [16]. Embedding these guidelines into NDT practice ensures that analytical objectives are balanced with the overarching responsibility to safeguard the cultural, historical, and symbolic value of heritage objects. Against this ethical and procedural backdrop, this review examines recent advances in NDTs, classifying them into three major categories and discussing their applications, advantages, and limitations in the context of CH conservation.
In this review, we classify the latest progress in NDTs for CH characterization into three major categories: spectrum-based, X-ray-based, and digital-based techniques. The methodological evolution from molecular-level analytical detection to AI-driven digital diagnosis is particularly emphasized, reflecting the expanding role of NDTs in cultural heritage science. The representative methods in each category are discussed according to their principles, applications, advantages, and limitations. By drawing on case studies of different material categories (such as manuscripts, paintings, textiles, architecture, and archeological relics), we emphasized the practicability of each technique in revealing the composition, structure, and context information while maintaining the integrity of cultural relics. In addition, we also discussed how the integration of portable instruments, multimodal data fusion, and artificial intelligence can reshape heritage diagnosis, providing new ways for sustainable, data-driven protection strategies. To provide a concise comparative overview and enhance the didactic clarity of the methodological framework, we summarized the techniques discussed above in Table 1.

2. Non-Destructive Techniques (NDTs)

2.1. Spectroscopy-Based Techniques

Spectroscopy techniques have been widely used in cultural heritage (CH) research because of their ability to reveal chemical and molecular information with low requirement for samples. Among them, Fourier transform infrared (FTIR), Raman, and nuclear magnetic resonance (NMR) spectroscopy show excellent performance in the non-destructive analysis of a variety of CH materials.

2.1.1. FTIR Spectroscopy

FTIR spectroscopy characterizes molecular vibrations through infrared radiation absorption, providing chemical fingerprints of organic and inorganic materials [17,18,19,20]. This technique excels in identifying compounds across solid, liquid, and gaseous states, particularly organic components (e.g., proteins, polysaccharides, and binders) in CH contexts. Three common operation modes, including transmission, external reflection, and attenuated total reflectance (ATR), usually consider three main regions, including near-infrared (NIR, 12,500–4000 cm−1), mid-infrared (MIR, 4000–400 cm−1), and far-infrared (FIR, 400–50 cm−1) [21]. While the transmission mode may require micro-sampling (<1 mg), FTIR is fundamentally non-destructive, as it preserves molecular structures and chemical bonds, causes zero chemical alteration during analysis, and enables the re-analysis of prepared samples.
Key applications are systematized in Table 2, with representative studies detailed below. For paper- and cellulose-based materials, FTIR identifies cellulose degradation markers (C-O-C stretch at 1160 cm−1) and sizing agents (protein amide I/II bands). Liu et al. [12] demonstrated its capacity to chronologically classify papers. Medieval rag paper exhibits starch/gelatin signatures (peaks at 1650 and 1540 cm−1); industrial-era paper shows lignin peaks (at 1510 and 1600 cm−1), indicating a risk of acid hydrolysis; and modern paper contains CaCO3 filler (peaks at 875 and 1420 cm−1) for acid stabilization. This chronological analysis informs preservation strategies [22]. In the study of textiles and natural fibers, the differentiation between proteins and polysaccharides elucidates the origins of materials. A pillowcase, dating from the 19th to 20th centuries, is composed of two components: the base textile fiber and the dyed fiber. ATR-FTIR spectroscopic analysis confirms that the base textile fiber is flax, whereas the dyed fiber is wool. Wool is more readily identifiable via ATR-FTIR, thanks to its distinctive amide I and amide II bands, along with other prominent bands like the 1390 cm−1 band (attributed to CH3 symmetrical deformation) and the 1055 cm−1 band (resulting from antisymmetric C–O–C stretching and C–N stretching), as illustrated in Figure 1a [23]. Silk, characterized by its β-sheet fibroin, exhibits peaks at 1625 and 1520 cm−1 [24]. Wool, composed of α-helical keratin, shows peaks at 1650 and 1545 cm−1 [25]. Plant fibers, such as flax and hemp, can be distinguished by shifts in cellulose crystallinity, with peaks at 1335 and 897 cm−1. Additionally, ATR mapping reveals the heterogeneity of degradation. Leather displays collagen triple helix signatures, with amide I characterized by C=O stretching (1630–1660 cm−1), amide II by N-H bending coupled with C-N stretching (1540–1560 cm−1), and amide III by C-N stretching plus N-H bending (1230–1300 cm−1) [26,27,28]. In the context of paints and coatings, the formation of metal carboxylates, also known as soap aggregates, is a significant deterioration mechanism. This process is monitored through the asymmetric stretch of COO−. For zinc soaps, the range is 1540–1590 cm−1 [29], while for lead soaps, it is 1500–1530 cm−1. Case studies on Van Gogh’s works demonstrate how Fourier transform infrared spectroscopy (FTIR) can distinguish between zinc palmitate at 1542 cm−1 and amorphous carboxylates at 1575 cm−1 [30], thereby facilitating targeted conservation treatments. For stone, ceramics, and polymers, cross-material diagnostics utilize the versatility of FTIR spectroscopy: Stone monuments reveal bioweathering through the presence of oxalate weddellite (peaks at 1320 and 780 cm−1). In archeological ceramics, silicate networks (peaks between 1000 and 1100 cm−1) indicate firing conditions [31]. For polymers, portable ATR-FTIR can differentiate between PVC (peaks at 1420 and 1330 cm−1) and cellulose acetate (peak at 1740 cm−1) in 3D objects without the need for sampling. For integrated methodologies, as established in foundational studies. FTIR synergizes with X-ray fluorescence (XRF) for pigment/metal ion correlation and Raman spectroscopy for complementary molecular specificity [32,33]. This multi-technique framework delivers holistic material characterization across easel paintings, wall paintings, and composite artifacts.

2.1.2. Raman Spectroscopy

Raman spectroscopy analyzes molecular vibrations through frequency-shifted scattered light [34]. When monochromatic light interacts with a sample, most scattered light is elastic (Rayleigh scattering), but a small fraction undergoes inelastic Raman scattering due to photon–molecule energy exchange. The resulting frequency shifts (Raman shifts) directly correspond to molecular rotational/vibrational energy levels, providing chemical bond and structural information [35]. Critically, this technique is inherently non-destructive, its low-energy incident light causes no sample alteration, and molecular polarizability-dependent signal acquisition typically requires no preparation. These advantages, combined with technological advances in portability since the first mobile instrument for CH applications was introduced in 2004 [36,37], make Raman spectroscopy particularly powerful for in situ CH diagnostics. For instance, it excels in pigment analysis by accessing low-wavenumber regions (<500 cm−1), a range challenging for conventional far-infrared spectrometers.
Tournié et al. reported the preliminary results of an in situ Raman spectroscopy analysis of San rock paintings in South Africa [38]. Their study had three main aspects: evaluating the performance of the Raman instrument under challenging conditions, analyzing the pigments and components used, and assessing the contributions of in situ analysis. However, the total weight of the instrument is 60 kg, and a lengthy assembly process is required before use. Hernanz et al. used in situ micro-Raman spectroscopy (μ-RS) and SEM-EDX to analyze the pigments in rock paintings in open-air rock shelters and detected the presence of calcite, dolomite, hematite, and amorphous carbon [39]. Prieto-Taboada et al. used non-destructive X-ray fluorescence spectroscopy, infrared spectroscopy, and Raman spectroscopy to analyze the composition of blue pigments in the murals of the Ariadne Palace (Pompeii, Italy) and the cause of pigment decay [40]. The presence of turquoise was successfully detected in the lower layer of Egyptian blue. Ma et al. used portable Raman spectroscopy and energy-dispersive X-ray fluorescence spectroscopy (ED-XRF) to identify pigments on the murals of Yanshan Temple in Fanshi County [41]. Therefore, it not only helps in understanding the social and cultural background of the time through pigments but also provides pioneering significance for the restoration and preservation of the murals. Mosca et al. used an in situ Raman device based on remote probes to analyze pigments used on badges or manuscript symbols [42]. It was found that the low flow rate on the surface of the sample can prevent overheating and possible damage to the sample.
Raman spectroscopy is a powerful tool for detecting pigment composition in rocks, murals, manuscripts, and sculptures, which helps us analyze the cultural heritages of glass, ceramics, and porcelain [43]. Colomban et al. used a portable Raman spectrometer equipped with two 532 nm Nd/YAG Ventus lasers to identify carnelian and glass trading beads and compared them with beads collected from different locations in southern Africa and those excavated on Mayotte Island [44]. They conducted an engineering analysis of the first batch of ceramics produced in Europe by combining Raman spectroscopy and XRF and demonstrated the effectiveness of Raman scattering in detecting secondary phases [45].
In recent years, surface-enhanced Raman spectroscopy (SERS) technology has made great progress and gradually gained widespread applications [46]. SERS is a technique with a surface enhancement effect used to amplify Raman signals. The surface enhancement effect is usually achieved by placing the sample of interest on a substrate with a special surface structure or coating that induces local light enhancement, thereby amplifying the Raman signals of molecules adsorbed on the sample surface. This amplification effect could significantly increase the intensity of Raman signals, thereby improving detection sensitivity. SERS is increasingly being used to identify organic dyes in CH [47]. Compared with ordinary Raman spectroscopy, SERS requires micro-sampling (samples ranging from 20 to 100 microns) and provides stronger and clearer resolved spectra for the identification of CH. Silvia et al. made significant contributions by using SERS to study natural organic dyes, especially dyes in dyed textiles, including dragon blood, sandalwood, annatto, safflower yellow, red, old fustic, gamboge, catechu, kamala, aloe, and sap green [48]. Most published examples are red anthraquinone dyes, such as carmine and madder. Notably, Leona demonstrated that surface-enhanced SERS, an optimized form of SERS using resonance excitation (488 nm), coupled with a monodisperse microwave-synthesized silver colloid and non-extractive HF vapor pretreatment, enables the ultra-sensitive detection of anthraquinone lakes embedded in binding media [49]. Using this method, Leona successfully identified the earliest known occurrence of madder lake on an Egyptian polychrome leather fragment from the Middle Kingdom (~2000 BC), revealing the sophisticated chemical knowledge in ancient dye technology and highlighting the potential of SERS-based methods for the non-invasive analysis of organic pigments in polychrome artworks (as shown in Figure 1b).
Heritage 08 00381 g001
Figure 1. (a) The application of FTIR in the identification of textiles [23], and (b) SERS analysis of madder lake in a painted leather quiver fragment from Middle Kingdom Egypt (2124–1981 BC), accession No. 28.3.5, with dimensions of 11 cm in height and 13 cm in width [49].
Figure 1. (a) The application of FTIR in the identification of textiles [23], and (b) SERS analysis of madder lake in a painted leather quiver fragment from Middle Kingdom Egypt (2124–1981 BC), accession No. 28.3.5, with dimensions of 11 cm in height and 13 cm in width [49].
Heritage 08 00381 g001

2.1.3. NMR Spectroscopy and Relaxometry

Nuclear magnetic resonance (NMR) spectroscopy provides detailed molecular-level information non-destructively and non-invasively, avoiding mechanical or chemical damage to artifacts [50]. Its sensitivity to the local chemical environment makes it valuable for analyzing sensitive CH materials [51].
The application of NMR in CH has evolved from high-field laboratory NMR spectroscopy (requiring mg-g samples) to portable low-field instruments developed over the past two decades. Crucially, portable devices primarily perform NMR relaxometry, not high-resolution spectroscopy. They measure nuclear relaxation times (T1 and T2), providing information on molecular dynamics and key physico-chemical properties like porosity and moisture without yielding detailed molecular spectra. This distinction is important: relaxometry offers non-invasive bulk property data, while spectroscopy provides structural detail typically requiring micro-sampling. This evolution has significantly expanded NMR’s applicability and reduced invasiveness [52,53].
Portable NMR relaxometry enables on-site, non-destructive analysis [54,55]. A critical aspect is depth-resolved analysis, achievable using different probes designed for specific measurement depths. This is particularly advantageous for large or immovable objects (murals, statues, architecture). A portable NMR device was used to detect the moisture content in the mural “Saint Clement at Mass and the Legend of Sisinnius” in the Basilica of St. Clement in Rome, as shown in Figure 2a [56]. It is crucial for identifying high-risk areas and targeted preservation. Additionally, portable NMR is critical for stone deterioration (pollutant ingress). T2 relaxation correlates strongly with pore structure, validated against mercury intrusion porosimetry (MIP) [57]. However, portable relaxometry operates at lower fields than laboratory spectroscopy, inherently limiting spectral resolution and precluding detailed molecular analysis. It remains highly valuable for the rapid in situ assessment of immovable objects.
High-field NMR spectroscopy provides unique insights into material composition, structure, and dynamics [58]. Solid-state NMR excels for insoluble materials (leather, stone). For instance, characteristic chemical shifts (10–70 ppm) aid amino acid identification in peptides [59]. Zhang et al. monitored leather collagen degradation using specific amino acid indicators (Hyp, Gly, Pro) [60]. High-resolution analysis typically requires mg-level powdered samples.
MRI generates non-destructive 3D images of internal structures, revealing hidden features or damage. Micro-MRI (mMRI) identified archeological reeds from Tutankhamun’s tomb (Figure 2b [61]), distinguishing cane species without mechanical sectioning [61]. The sample amount depends on the type of target molecules and the sensitivity of the instrument, but it generally ranges from several milligrams to a few grams, ensuring sufficient signal detection while minimizing invasiveness.
Heritage 08 00381 g002
Figure 2. (a) A portable NMR device used to calculate the moisture content in the mural “Saint Clement at Mass and the Legend of Sisinnius” in the Basilica of St. Clement in Rome [56], and (b) the micro-magnetic resonance imaging (mMRI) of archeological reeds found in Tutankhamun’s tomb [61].
Figure 2. (a) A portable NMR device used to calculate the moisture content in the mural “Saint Clement at Mass and the Legend of Sisinnius” in the Basilica of St. Clement in Rome [56], and (b) the micro-magnetic resonance imaging (mMRI) of archeological reeds found in Tutankhamun’s tomb [61].
Heritage 08 00381 g002

2.2. X-Ray-Based Techniques

2.2.1. Conventional XRD and XRF

X-rays could generate diffraction patterns with ordered structural arrangements and excite the elements in the sample to produce characteristic fluorescence [62]. So X-ray diffraction (XRD) and X-ray fluorescence (XRF) techniques have been developed [63].
XRD is a popular non-destructive technique for analyzing solid crystalline materials. The scattering of X-rays at specific angles on the lattice planes of a sample is calculated and converted into its crystal structure. Each type of crystalline compound has its unique XRD pattern, which could serve as a comparative feature “fingerprint”. XRF is a powerful analytical technique used to determine the elemental composition of materials. It involves exposing the sample to high-energy X-rays, which can cause the atoms in the sample to emit secondary (or fluorescent) X-rays. The energy and intensity of these emitted X-rays are characteristic of the specific elements in the material, which can be accurately identified and quantified. XRF is widely used in various fields such as geology, archeology, art preservation, and material science because it can quickly and accurately analyze various elements from heavy metals to light elements [64]. Its advantages include minimal sample preparation, fast analysis time, and the ability to analyze both solid and liquid samples. In addition, portable XRF (pXRF) devices could perform on situ analysis, making on-site investigations possible [65].
The application of XRD and XRF plays a crucial role in the color analysis of ancient artifacts. A research project was conducted on an extraordinary leather painting in the Alhambra’s Hall of the Kings in Granada, Spain. The painting depicts various kings, a fountain of youth, and a queen playing chess [66]. In the study, 11 and 14 points were collected from the sections depicting the kings and fountain of youth sections, respectively, as shown in Figure 3a [66]. They used a portable device that combined XRD and XRF techniques to evaluate the color compositions and mineralogical phases of pigments without sampling from the artwork. The results showed that reds were made from cinnabar and/or hematite, flesh was a mixture of cinnabar and hydrocerussite, greens were made from malachite and orpiment, blues were made from lazurite or azurite, and ocher was made from clay. Through XRD, red lead and gold and tin were found in the red ocher and oranges hues, while strontium was related to gypsum preparation. The presence of toxic pigments in gypsum paint prompted repairers to take special preventative measures. Therefore, in the review, some color information obtained by XRD and XRF is summarized in Figure 3b. J. Daniel Martin-Ramos et al. used XRD techniques to examine the pigments in the painting “Little Madonna of Foligno” [67]. They found that the blue was made of lapis lazuli and some chalcopyrite, while the red was made of cinnabar. The use of these expensive natural pigments suggested that this painting should be commissioned by high-leveled individuals such as nobles, merchants, or clergy. In addition, all shades of blue contained lapis lazuli, a pigment commonly used by Renaissance painters for high-profile clients, except for one spot containing the untimely Prussian blue, indicating that the area might have been restored after the 18th century. Luo et al. combined XRF with other techniques to identify leather material residues covered with red pigments on a Chinese bronze sword from the 6th century B.C. [68]. The results revealed that the red pigment present on the leather was identified as cinnabar (HgS). Compared to the calcium and iron in the surrounding tomb soil, higher levels of calcium (in the form of CaCO3) and iron (in the form of Fe2O3) were found. These findings suggested the use of lime for hair removal and degreasing, as well as tanning agents composed of iron–aluminum compound salts. Furthermore, it could be inferred that this technique of using iron–aluminum salt for tanning had been developed in the mid to late Spring and Autumn Period in China. While pXRF enables rapid, on-site elemental screening with minimal sample preparation and broad elemental coverage, its lower sensitivity for trace elements, reduced accuracy for very light elements, and susceptibility to matrix effects often limit its application to qualitative or semi-quantitative analyses. These characteristics make pXRF an effective preliminary screening tool, with high-sensitivity laboratory techniques (e.g., total reflection XRF) providing precise follow-up data.

2.2.2. Total Reflection XRF (TRXRF)

Building upon conventional XRF, total reflection XRF (TRXRF) is optimized for ultra-trace and ultra-thin surface characterization. By operating at extremely low incident angles (<0.1°), TRXRF achieves total external reflection of the primary X-ray beam. This restricts penetration depth to a few nanometers, drastically reducing matrix effects while enhancing surface sensitivity. Therefore, TRXRF attains detection limits in the mg/g (solids) and ng/g range (liquids) [69]. Compared with pXRF, TRXRF typically offers one to two orders of magnitude lower detection limits, requires more controlled sample preparation (e.g., deposition on a flat, smooth substrate), and delivers higher analytical precision for surface and trace element quantification. While pXRF excels in rapid, non-destructive, on-site elemental screening, TRXRF is ideal for detailed laboratory investigations. This synergy establishes their complementary utility in CH studies.
TRXRF has special value in analyzing archeological ceramics, especially in trace element analysis and provenance research. In a study of Neolithic pottery from the Lake Baikal region, TRXRF was used to obtain multi-elemental fingerprints, which were then analyzed using principal component analysis (PCA), k-means clustering (a statistical clustering technique that minimizes intra-cluster variance), and support vector machines (SVMs) [70]. The results enabled the effective classification of the ceramic samples based on their sources and types, confirming the practicality of this technology in reconstructing ancient production and distribution networks. Another investigation evaluated the semi-quantitative ability of TRXRF on archeological ceramics [71]. By comparing two sample preparation methods, solid sedimentation and chemical homogenization, the study evaluated the impact of preparation method on elemental measurements and compared the results with those obtained through instrumental neutron activation analysis (INAA). TRXRF was found to provide a good balance between analytical precision and minimal sample consumption, indicating its suitability not only for classification but also for reproducible semi-quantitative compositional studies. In addition to ceramics, TRXRF has also demonstrated wide applicability to various cultural heritage materials. In the context of archeological surveys in Argentina, this technique was used to analyze pigments in rock art and surface residues from underwater artifacts [72]. Its ability to detect ultra-trace elements without sampling was particularly advantageous in situations where storage restrictions limit any form of invasive analysis. In summary, TRXRF has become a forward-looking analytical technique in cultural heritage science, particularly in the context of modern research where conservation and non-destructiveness are crucial. Owing to its high sensitivity, strong surface selectivity, and minimal sample requirements, TRXRF provides a practical balance between analytical accuracy and protecting artifact integrity.

2.2.3. Spatially Resolved X-Ray Techniques

Another extension of XRD is micro-XRD (μ-XRD), which can be used to study local restoration, corrosion, or other subtle changes in artifacts. It is particularly suitable for situations that require high-resolution structural information at the microscopic scale. For example, Franquelo et al. analyzed the cross-section of decorated leather samples using μ-XRD and found that the original pigment layer on the leather was composed of gypsum [69]. In contrast, macro-XRD (MA-XRD) could provide valuable insights into the volume characteristics and overall crystallographic composition of the object when artifacts are large or a more comprehensive structural overview is needed. In order to adapt large or irregularly shaped surfaces, a combined MA-XRD/MA-XRF system was developed, which featured a micro-focal copper anode source, multiple capillary focusing optic, and dual detectors, capable of simultaneously capturing XRD patterns and XRF spectra in the same region. This configuration enables a comprehensive structural and elemental analysis of complex artworks without the need for physical sampling [73].
X-ray-based methods, including XRD, XRF, TRXRF, and spatially resolved variants (μ-XRD, MA-XRD/MA-XRF), enable the non-destructive structural and elemental characterization of CH materials with high specificity. XRD provides crystalline-phase “fingerprints” for mineral identification, while XRF delivers rapid, multi-element analysis with minimal sample preparation and broad applicability to both solids and liquids. TRXRF offers ultra-trace detection and strong surface selectivity, making it valuable for provenance studies and thin-layer analysis. Spatially resolved systems allow for the micro- to macro-scale mapping of structural and compositional features in situ, even on large or irregular objects. However, limitations include lower sensitivity for light elements (especially in pXRF), matrix effects affecting quantification, and generally lower precision for portable systems compared to laboratory instruments. TRXRF requires controlled sample preparation and is limited to laboratory use, while high-resolution mapping systems involve complex instrumentation and high operational costs.

2.3. Digital-Based Techniques

The latest advances in digital technologies have transformed cultural heritage research by enabling non-invasive, high-resolution, and data-driven investigations. These techniques can be roughly divided into four functional groups: digital imaging and visualization, volume and structure imaging, data fusion and computational processing, and AI-assisted diagnosis and virtual restoration. Each category reflects different dimensions of digital workflow, from data collection and structural modeling to automatic interpretation, highlighting their unique contributions and potential for comprehensive heritage preservation strategies. The following sections provides an overview of their principles and applications and representative case studies.

2.3.1. Digital Imaging and Visualization

Digital imaging techniques capture two-dimensional visual or spectral information from artifact surfaces, enabling a non-contact analysis of surface morphology, materials, and deterioration. High-resolution photography, UV/IR imaging, multispectral/hyperspectral imaging (MSI/HSI), and infrared thermography (IRT) are widely employed to identify faded patterns, pigment distributions, and sub-surface features. Hyperspectral imaging (HSI) has proven particularly valuable for non-invasive surface analysis, with Liang et al. demonstrating its narrow-band spectral resolution for pigment identification [74], and advanced techniques like PCA and SAM distinguish overlapping materials in sketch paintings [75]. Recent extensions integrate HSI with convolutional neural networks (CNNs) to automate defect recognition, such as identifying degraded Egyptian hieroglyphs [76].
Complementing surface analysis, infrared thermography (IRT) detects radiation in the 700 nm to 1 mm range to evaluate internal defects and moisture in architectural, sculptural, paper-based, and painted heritage [77]. Passive IRT captures natural heat emissions, applied effectively to assess the thermal compatibility of repaired stones in historic buildings like Kurucesme Han [76] and detect delamination/humidity in medieval murals [77]. Active IRT employs controlled thermal stimulation, with pulse thermography (PT) using short-duration heating (e.g., 10 μs flashes) and dual-band cameras (MWIR: 4.4–5.2 μm; LWIR: 7.8–8.8 μm) to quantify metal corrosion in cast iron and locate embedded elements in bronze sculptures [78]. Recent innovations include pulse compression thermography (PCT), which combines encoded excitation (e.g., chirp signals) followed by a principal component analysis of the thermal response sequence to amplify low-contrast defects. This approach has become critical for panel painting consolidation assessments and marquetry delamination mapping, where processed thermal sequences reveal adhesive failures [79]. In the study of the 15th-century detached wall painting attributed to Antonio del Massaro, the combined application of MSI and PCT exemplifies multimodal synergy [80].
The reliability of IRT findings is strengthened through cross-validation. A study on automated moisture detection in heritage walls used 5-fold cross-validation to evaluate CNN models for infrared thermal images, ensuring result stability. For moisture detection, the model achieved an average accuracy of 97.26% with consistent precision, recall, and F1-score, validating IRT’s robustness in assessing heritage structures [81].

2.3.2. Volumetric and Structural Imaging

Volumetric and structural imaging techniques provide three-dimensional internal or external reconstructions of heritage objects. These methods are critical for evaluating structural integrity, fabrication methods, and hidden features. Common methods include computed tomography (CT), micro-CT (μ-CT), neutron tomography, photogrammetry, structure-from-motion (SfM), and LiDAR-based scanning, each providing different resolutions and penetration depths, suitable for different material types and artifact scales [82,83]. There are currently three types of radiation, including X-rays, neutrons, and gamma radiation, with X-rays being the primary choice. X-ray CT can be classified based on its resolution, with industrial CT typically having a resolution between 5 and 150 μm. In contrast, μ-CT provides higher resolution ranging from 0.5 to 5 μm. In addition, nano-CT could provide even finer resolutions of less than 0.5 μm [84]. Compared with nano-CT, micro-CT is widely used in the study of CH because it can capture complex details in artifacts while maintaining relatively inexpensive contours [85]. For multi-layered organic composites such as ancient marquetry, μ-CT (220 kV, voxel size 116.47 μm) enables the precise quantification of structural degradation. As shown in Figure 4a, a study of a 208 × 212 mm marquetry sample revealed the following: (1) layer-specific metrology with a decorative tessellatum thickness of 1.2–1.7 mm and wooden support of 13.5–14.5 mm and (2) defect characterization including wood cracks, pores, delaminations at adhesive interfaces, and mineral inclusions in organic tesserae (e.g., horn, mother-of-pearl) [86]. This case underscores μ-CT’s unique capacity to resolve heterogeneous organic layers, where conventional surface techniques fail [87,88]. Through CT technology, the hidden internal details of artifacts could be revealed, which is particularly evident in the context of paper artifacts, such as paintings and ancient manuscripts. For example, one of the most famous paintings in history, the Mona Lisa, was found to have a different face beneath its surface. According to rumors, the hidden face reveals the true model of this painting, which is Duchess Isabella, rather than a woman named Mona Lisa Gherardini, (Figure 4b) [89]. In 2017, researchers from Italy revealed that a painting called Ecstasy of Saint Teresa had surprising hidden details in a CT scan. Beneath the surface of the painting, there is an angel whose arrow points towards the heart of the Saint [90]. CT is also used to study ancient books, as any attempt to read text carries an irreversible risk of destruction. Using CT, researchers are able to read the content without physical contact. In addition, CT can be used to examine complex component objects. Bulcke et al. used a laboratory-based X-ray CT system to analyze various wooden artifacts, including a cello, acoustic guitar, violin, bow, and pipa. The instrument slices obtained from the CT scan were then processed to create a 3D model, in which internal structures were clearly observed. From the cross-sectional image, it was observed that the contrast between organic components (such as the guitar body) and the background was low, while the contrast between metal components (such as strings) was high. However, by using neutron CT, the problem of valuable information being lost due to specific artifacts caused by low object contrast can be effectively addressed. In a study conducted in Switzerland, researchers examined a short sword using neutron and X-ray CT [91]. The wooden part of the handle showed higher contrast in neutron CT, while the contrast recorded in X-ray CT was significantly lower. Therefore, neutron CT could serve as be a supplementary technique to X-ray CT. Copper and iron metals have long been used for mental CH. Compared to organic materials, artifacts made from them contain heavy atoms, resulting in lower X-ray CT transmittance and significantly improved contrast. However, excessive contrast between the image and background can also mask the details of CH. In order to increase the transmittance of X-rays, high-energy X-ray CT (170–200 keV) is usually used [92]. In a study conducted in Brazil, researchers analyzed copper coins and buttons using X-ray CT with a scanning voltage of 130 kV. The crust covering the artifacts was removed in micro-CT, and the hidden textures became visible in the 3D model created using CT images (Figure 4c) [93].

2.3.3. Data Fusion and Computational Processing

With the increasing popularity of multimodal datasets, data fusion and processing techniques are crucial for integrating and enhancing the interpretability of digital heritage information. This includes spectral–spatial image registration, 3D model–spectral image integration, virtual staining, color normalization, and texture enhancement.
The scope of data fusion extends to integrating information from advanced micro-analytical and large-scale facility techniques, which present unique processing challenges due to heterogeneous data dimensionality and spatial resolutions. For elemental and molecular characterization at micro-scales, techniques like laser-induced breakdown spectroscopy (LIBS) [94,95] and laser ablation inductively coupled plasma mass spectrometry (LA-ICP-MS) [96,97] generate high-dimensional spectral maps and elemental distribution profiles. Optical coherence tomography (OCT) [98] provides cross-sectional, sub-surface structural information. Large-scale synchrotron and ion beam facilities (e.g., employing µXRF, µXRD, PIXE, RBS, EXAFS) deliver unparalleled spatial resolution for elemental speciation, chemical state mapping, and crystalline structure analysis, often generating massive, multimodal datasets per sample region. Additionally, ion beam analysis (IBA) techniques, including proton-induced X-ray emission (PIXE), Rutherford backscattering spectrometry (RBS), and particle-induced gamma-ray emission (PIGE), provide complementary elemental depth profiles at micron scales but require vacuum conditions and specialized facilities.
Advanced algorithms, including those specifically designed for handling high-dimensional spectral data (e.g., from hyperspectral imaging encompassing UV-Vis-NIR ranges) and correlating multi-scale, multi-physics datasets (e.g., co-registering µXRF maps with OCT volumes or 3D surface models), enable the collaborative visualization and interpretation of diverse sources. This enhances material classification and damage assessment, as demonstrated at the Palacio de Colomina in Valencia where multi-source fusion reconstructed architectural evolution and supported damage tracking [99]. Here, a multispectral voxel fusion technique combined with self-organizing maps (SOMs) enhanced structural evaluation [100].
Interdisciplinary applications further showcase these advantages: At the Church of San Michele Arcangelo (Padula, Italy), multi-sensor integration diagnosed surface instability in mural-covered vaults [101]. Critically, integrating OCT data could reveal sub-surface delamination invisible to surface sensors, while fusing µXRF/LIBS elemental maps with hyperspectral imaging establishes direct links between composition and spectral signatures, significantly enhancing alteration product mapping. These examples collectively demonstrate that integrating digital data streams plays an increasingly important role in improving document accuracy, structural diagnosis, and conservation decision in heritage science.

2.3.4. AI-Assisted Diagnosis and Virtual Restoration

From a methodological perspective, AI approaches in CH can be grouped into (i) supervised learning for classification, detection, and segmentation (e.g., CNN/U-Net/Mask R-CNN, transformer-based models) when labeled data are available; (ii) unsupervised or self-supervised learning for clustering, anomaly detection, and representation learning (e.g., PCA/NMF, k-means/SOM, autoencoders), which is useful under scarce annotations; (iii) generative modeling (e.g., GANs, diffusion) for inpainting, 3D completion, and virtual restoration; and (iv) semi-supervised and transfer learning to leverage limited, domain-specific datasets. This taxonomy provides a consistent lens through which the following applications can be interpreted.
Artificial intelligence (AI) is playing an increasingly prominent role in cultural heritage science, particularly in the fields of damage diagnosis and digital restoration. On the supervised side, high-performance detectors and segmenters have been deployed for surface condition assessment, for example, Mask R-CNN detecting gold foil shedding on the Dazu Thousand-Hand Bodhisattva with high average precision and YOLO-family models identifying wall/roof damage in traditional buildings in Zhejiang, China. Crucially, supervised approaches now extend to cost-effective pigment classification using conventional RGB imagery. Genc et al. [102] demonstrated that ResNet50/VGG19 models trained on 8332 images of historical pigments (ochres, Egyptian blue, ultramarine) under varied illumination (3415–6512 K) achieved 99–100% accuracy in identifying pure pigments on artworks like Bosch’s The Haywain and Egyptian mummy masks. This paradigm shift eliminates dependency on specialized sensors, enabling non-destructive analysis via smartphone photography alone. For generative modeling, recent studies have demonstrated GAN- or diffusion-based inpainting for mural restoration and 2D reconstructions (e.g., Roman coins), as well as deep learning pipelines for the 3D virtual restoration of degraded artifacts. These approaches combine semantic segmentation, learned feature representations, and generative modules to recover missing content while preserving stylistic plausibility for exhibition and analysis. Beyond single-site casework, AI has also been applied at the landscape scale using aerial/orthophoto data to detect maintenance issues (e.g., vegetation overgrowth, damaged conduits) at archeological sites such as Pompeii, illustrating the feasibility of non-intrusive, large-area diagnostics.
Dataset standardization is pivotal for comparability and reproducibility in AI-assisted diagnostics. Practical requirements include harmonized acquisition protocols (illumination/geometry, white–dark calibration, instrument spectral response), standardized file formats and rich metadata, and consistent annotation guidelines with inter-annotator agreement reporting. Transparent data splits, cross-site/instrument validation, and matrix-matched references or phantoms help establish fair benchmarks across laboratories. Community frameworks (e.g., IIIF for interoperable image delivery; CIDOC-CRM as an ISO ontology for semantic integration) and EU initiatives toward a common European data space for cultural heritage provide foundations for dataset interoperability and scaling.
Interpretability and uncertainty quantification are also key. Post hoc tools (e.g., saliency maps/Grad-CAM, feature attribution) and spectral/spatial attention visualization make decisions auditable by conservators, while physics- or rule-informed constraints improve plausibility. Reporting calibrated confidence (e.g., deep ensembles, Monte Carlo dropout) supports risk-aware deployment and human-in-the-loop verification. To provide a broader perspective, Marco et al. surveyed AI in heritage and categorized applications into supervised, unsupervised, and semi-supervised learning models, across tasks such as material classification, stylistic attribution, site prediction, and digital archiving, while highlighting challenges in standardized datasets, feature interpretability, and interdisciplinary collaboration [103]. To systematically summarize AI techniques in CH diagnostics, we categorized them based on learning paradigms and dataset characteristics (Table 3).
Digital-based methods (spanning imaging, volumetric scanning, data fusion, and AI-assisted workflows) enable the non-invasive, high-resolution, and multimodal documentation of cultural heritage. Digital imaging (e.g., HSI, MSI, IRT) excels at rapid, wide-area surface mapping, revealing pigment distributions, hidden patterns, and deterioration features without contact. Volumetric and structural imaging (e.g., CT, μ-CT, neutron tomography, photogrammetry) provides 3D internal and external reconstructions, supporting structural analysis, manufacturing studies, and non-invasive reading of concealed details. Data fusion and computational processing integrate heterogeneous datasets (spectral, spatial, structural) to improve interpretability, while AI-assisted diagnosis and virtual restoration enable automated damage detection, virtual reconstruction, and large-scale monitoring. Meanwhile, limitations include equipment cost, technical complexity, and data size/processing demands, which may restrict access for smaller institutions. Some volumetric methods require immobile, high-end instruments (e.g., neutron CT) or are constrained by object size and penetration depth. AI-based approaches depend on high-quality, standardized datasets and remain sensitive to annotation bias and model interpretability issues. Despite these constraints, digital-based techniques offer scalable, integrative solutions for both preventive conservation and research, especially when combined in multimodal workflows.

3. Conclusions and Prospect

The development of NDTs has fundamentally changed the landscape of cultural heritage science, creating a multidimensional toolkit spanning molecular precision (spectroscopy), structural granularity (X-ray), and immersive interpretation (digital imaging). Despite this transformation, critical challenges persist: the field remains hindered by fragmented multimodal integration due to underdeveloped unified platforms for structural–spectral–contextual synergy, compromised by the absence of universal data standards that limit cross-platform comparability and dataset interoperability and constrained by a pronounced field-to-lab translation gap where validation frameworks fail to capture real-world complexity [104].
To address these bottlenecks, future efforts must prioritize cross-technique calibration through inter-laboratory studies, heritage-specific calibration protocols, and certified reference materials (CRMs) to bridge sensitivity and matrix effect variations; advance AI-driven platform unification via deep learning and generative modeling for virtual restoration and predictive diagnostics, contingent on standardized heritage datasets; and develop scalable validation methodologies that reconcile controlled experiments with field variability. Only through the targeted resolution of these gaps can NDTs fully transcend their role as an analytical tool to become the cornerstone of heritage preservation paradigms, ensuring the material legacy of human history endures for future generations.

Author Contributions

M.Z. contributed to conceptualization, investigation, and writing—reviewing and editing. S.L. contributed to the literature review and summary. H.S. contributed to the literature review and summary. Z.B. contributed to the analysis and refinement of the section on X-ray-based techniques. M.G.A.K. was involved in visualization and validation. J.L. was responsible for visualization and validation. K.T. supervised this study, provided validation and resources, and secured funding. G.H. provided supervision and was involved in visualization and validation. All authors have read and agreed to the published version of the manuscript.

Funding

This work was funded by National Natural Science Foundation of China (No. 52373109, 52073262) and Science and Technology Department of Henan Province, China (No. 232102521017).

Data Availability Statement

No new data were generated in this study. The figures reproduced from previously published sources were used with proper permissions and copyright clearance.

Conflicts of Interest

The authors declare no conflicts of interest.

Abbreviations

The following abbreviations are used in this manuscript:
CHCultural heritage
NDTNon-destructive technique
XRFX-ray fluorescence
XRDX-ray diffraction (XRD)
ATRAttenuated total reflectance
μ-RSMicro-Raman spectroscopy
ED-XRFEnergy-dispersive X-ray fluorescence spectroscopy
SERSSurface-enhanced Raman spectroscopy
NMRNuclear magnetic resonance
MIPMercury intrusion porosimetry
MRIMagnetic resonance imaging
TRXRFTotal reflection XRF
MA-XRDMacro-XRD
μ-XRDMicro-XRD
IRTInfrared thermography
PTPulse thermography
HSIHyperspectral imaging
MSIMultispectral imaging
CTComputed tomography
μ-CTMicro-CT
SfMStructure-from-motion
AIArtificial intelligence
CNNConvolutional neural network
LIBSLaser-induced breakdown spectroscopy
LA-ICP-MSLaser ablation inductively coupled plasma mass spectrometry
OCTOptical coherence tomography
IBAIon beam analysis
PIXEProton-induced X-ray emission
RBSRutherford backscattering spectrometry
PIGEParticle-induced gamma-ray emission
CRMCertified reference material

References

  1. Jagielska-Burduk, A.; Pszczyński, M.; Stec, P. Cultural heritage education in UNESCO cultural conventions. Sustainability 2021, 13, 3548. [Google Scholar] [CrossRef]
  2. Muradov, M.; Kot, P.; Markiewicz, J.; Łapiński, S.; Tobiasz, A.; Onisk, K.; Shaw, A.; Hashim, K.; Zawieska, D.; Mohi-Ud-Din, G. Non-destructive system for in-wall moisture assessment of cultural heritage buildings. Meas 2022, 203, 111930. [Google Scholar]
  3. Forrest, C. An international perspective on sunken state vessels as underwater cultural heritage. Ocean. Dev. Int. Law 2003, 34, 41–57. [Google Scholar] [CrossRef]
  4. Cappa, F.; Piñar, G.; Brenner, S.; Frühmann, B.; Wetter, W.; Schreiner, M.; Engel, P.; Miklas, H.; Sterflinger, K. The kiev folia: An interdisciplinary approach to unravelling the past of an ancient slavonic manuscript. Int. Biodeterior. Biodegrad. 2022, 167, 105342. [Google Scholar] [CrossRef]
  5. Minhus, S.M.; Huie, L. The tendency of traditional costume at heritage festival for cultural revival. SAGE Open 2021, 11, 215824402110169. [Google Scholar] [CrossRef]
  6. Sallam, A.; Hemeda, S.; Toprak, M.S.; Muhammed, M.; Hassan, M.; Uheida, A. CT scanning and MATLAB calculations for preservation of coptic mural paintings in historic Egyptian monasteries. Sci. Rep. 2019, 9, 3903. [Google Scholar] [CrossRef]
  7. Mikayama, A.; Hokoi, S.; Ogura, D.; Okada, K.; Su, B. The effects of windblown sand on the deterioration of mural paintings in cave 285, in Mogao caves, Dunhuang. J. Build. Phys. 2018, 42, 652–671. [Google Scholar]
  8. Costantini, I.; Castro, K.; Madariaga, J.M. Portable and laboratory analytical instruments for the study of materials, techniques and environmental impacts in mediaeval mural paintings. Anal. Methods 2018, 10, 4854–4870. [Google Scholar] [CrossRef]
  9. Harth, A. X-ray fluorescence (XRF) on painted heritage objects: A review using topic modeling. Herit. Sci. 2024, 12, 17. [Google Scholar]
  10. Crina Anca Sandu, I.; de Sá, M.H.; Pereira, M.C. Ancient ‘gilded’ art objects from European cultural heritage: A review on different scales of characterization. Surf. Interface Anal. 2011, 43, 1134–1151. [Google Scholar] [CrossRef]
  11. Lehmann, E.H.; Vontobel, P.; Deschler-Erb, E.; Soares, M. Non-invasive studies of objects from cultural heritage. Nucl. Instrum. Methods Phys. Res. Sect. A 2005, 542, 68–75. [Google Scholar] [CrossRef]
  12. Liu, G.L.; Kazarian, S.G. Recent advances and applications to cultural heritage using ATR-FTIR spectroscopy and ATR-FTIR spectroscopic imaging. Analyst 2022, 147, 1777–1797. [Google Scholar] [CrossRef]
  13. Camara, A. International council of museums (icom): Code of ethics. In Encyclopedia of Global Archaeology; Springer International Publishing: Cham, Switzerland, 2020; pp. 5868–5872. [Google Scholar]
  14. EN 16085:2012; Conservation of Cultural Property—Methodology for Sampling from Materials of Cultural Property—General Rules. European Committee for Standardization (CEN): Brussels, Belgium, 2012.
  15. EN 16853:2017; Conservation of Cultural Heritage—Conservation Process—Decision Making, Planning and Implementation. European Committee for Standardization (CEN): Brussels, Belgium, 2017.
  16. EN 16095:2012; Conservation of Cultural Property—Condition Recording for Movable Cultural Heritage. European Committee for Standardization (CEN): Brussels, Belgium, 2012.
  17. Dutta, A. Fourier transform infrared spectroscopy. Spectrochim. Acta A Mol. Biomol. Spectrosc. 2017, 10, 73–93. [Google Scholar]
  18. Campanella, B.; Palleschi, V.; Legnaioli, S. Introduction to vibrational spectroscopies. ChemTexts 2021, 7, 5. [Google Scholar] [CrossRef]
  19. Zhang, W.B. Review on analysis of biodiesel with infrared spectroscopy. Renew. Sustain. Energy Rev. 2012, 16, 6048–6058. [Google Scholar] [CrossRef]
  20. Goda, K.; Jalali, B. Dispersive Fourier transformation for fast continuous single-shot measurements. Nat. Photonics 2013, 7, 102–112. [Google Scholar] [CrossRef]
  21. Pasquini, C. Near infrared spectroscopy: A mature analytical technique with new perspectives—A review. Anal. Chim. Acta 2018, 1026, 8–36. [Google Scholar] [CrossRef] [PubMed]
  22. Librando, V.; Minniti, Z.; Lorusso, S. Ancient and modern paper characterization by FTIR and micro-Raman spectroscopy. Conserv. Sci. Cult. Herit. 2011, 11, 249–268. [Google Scholar]
  23. Raditoiu, V.; Chican, I.E.; Raditoiu, A.; Fierascu, I.R.; Fierascu, C.; Fotea, P. VR Technologies in Cultural Heritage; Duguleană, M., Carrozzino, M., Gams, M., Tanea, I., Eds.; Springer: Cham, Switzerland, 2019; pp. 3–9. [Google Scholar]
  24. Kozłowski, R.M.; Mackiewicz-Talarczyk, M. 1A-Introduction to natural textile fibres. In Handbook of Natural Fibres; Kozłowski, R.M., Mackiewicz-Talarczyk, M., Eds.; Elsevier: Amsterdam, The Netherlands, 2020; Volume 1, pp. 1–13. [Google Scholar]
  25. Vasconcelos, A.; Freddi, G.; Cavaco-Paulo, A. Biodegradable materials based on silk fibroin and keratin. Biomacromolecules 2008, 9, 1299–1305. [Google Scholar] [CrossRef]
  26. Zhang, M.; Ba, Z.; Liu, J.; Lei, Y.; Kaya, M.; Tang, K. Rapid and non-destructive identification of the leather artifacts using FTIR-PCA. J. Cult. Herit. 2025, 72, 150–159. [Google Scholar] [CrossRef]
  27. Liu, S.; Zhang, M.; Liu, J.; Lei, Y.; Kaya, M.G.A.; Tang, K. Long-term antioxidant and ultraviolet light shielding gelatin films for the preservation of leather artifacts. Int. J. Biol. Macromol. 2025, 291, 138981. [Google Scholar]
  28. Zhang, M.; Liu, J.; Li, B.; Lei, Y.; Albu Kaya, M.G.; Zhang, X.; Tang, K. Revealing the secrets of a lacquered leather artifact through molecular fingerprints. npj Herit. Sci. 2025, 13, 148. [Google Scholar] [CrossRef]
  29. Hermans, J.J.; Keune, K.; van Loon, A.; Iedema, P.D. An infrared spectroscopic study of the nature of zinc carboxylates in oil paintings. J. Anal. At. Spectrom. 2015, 30, 1600–1608. [Google Scholar] [CrossRef]
  30. Derrick, M.R.; Stulik, D.; Landry, J.M. Infrared Spectroscopy in Conservation Science; Getty Conservation Institute: Los Angeles, CA, USA, 1999. [Google Scholar]
  31. Medeghini, L.; Mignardi, S.; De Vito, C.; Conte, A.M. Evaluation of a FTIR data pretreatment method for principal component analysis applied to archaeological ceramics. Microchem. J. 2016, 125, 224–229. [Google Scholar] [CrossRef]
  32. Bitossi, G.; Giorgi, R.; Mauro, M.; Salvadori, B.; Dei, L. Spectroscopic Techniques in Cultural Heritage Conservation: A Survey. Appl. Spectrosc. Rev. 2005, 40, 187–228. [Google Scholar] [CrossRef]
  33. Bell, J.; Nel, P.; Stuart, B. Non-invasive identification of polymers in cultural heritage collections: Evaluation, optimisation and application of portable FTIR (ATR and external reflectance) spectroscopy to three-dimensional polymer-based objects. Herit. Sci. 2019, 7, 95. [Google Scholar]
  34. Smith, E.; Dent, G. Modern Raman Spectroscopy: A Practical Approach; John Wiley & Sons: Hoboken, NJ, USA, 2019. [Google Scholar]
  35. Rousaki, A.; Vandenabeele, P. In situ Raman spectroscopy for cultural heritage studies. J. Raman Spectrosc. 2021, 52, 2178–2189. [Google Scholar] [CrossRef]
  36. Edwards, H.G.; Vandenabeele, P.; Colomban, P. Raman Spectroscopy in Cultural Heritage Preservation; Springer: Cham, Switzerland, 2023. [Google Scholar]
  37. Coccato, A.; Caggiani, M.C. An overview of principal components analysis approaches in Raman studies of cultural heritage materials. J. Raman Spectrosc. 2024, 55, 125–147. [Google Scholar]
  38. Tournié, A.; Prinsloo, L.C.; Paris, C.; Colomban, P.; Smith, B.J. The first in situ Raman spectroscopic study of san rock art in South Africa: Procedures and preliminary results. J. Raman Spectros. 2011, 42, 399–406. [Google Scholar]
  39. Hernanz, A.; Ruiz-López, J.F.; Madariaga, J.M.; Gavrilenko, E.; Maguregui, M.; Fdez-Ortiz de Vallejuelo, S.; Martínez-Arkarazo, I.; Alloza-Izquierdo, R.; Baldellou-Martínez, V.; Viñas-Vallverdú, R. Spectroscopic characterisation of crusts interstratified with prehistoric paintings preserved in open-air rock art shelters. J. Raman Spectrosc. 2014, 45, 1236–1243. [Google Scholar]
  40. Prieto-Taboada, N.; Fdez-Ortiz de Vallejuelo, S.; Santos, A.; Veneranda, M.; Castro, K.; Maguregui, M.; Morillas, H.; Arana, G.; Martellone, A.; de Nigris, B. Understanding the degradation of the blue colour in the wall paintings of Ariadne’s house (Pompeii, Italy) by non-destructive techniques. J. Raman Spectrosc. 2021, 52, 85–94. [Google Scholar] [CrossRef]
  41. Ma, J.; Li, Y.; Wu, F.; Wang, Y.; Guo, H.; He, X.; Zhang, X.; Wang, F. In situ analysis of pigments on the murals of Yanshan Temple in Fanshi County, Shanxi Province, China by portable Raman and XRF spectrometers. Eur. Phys. J. Plus 2023, 138, 48. [Google Scholar] [CrossRef]
  42. Mosca, S.; Frizzi, T.; Pontone, M.; Alberti, R.; Bombelli, L.; Capogrosso, V.; Nevin, A.; Valentini, G.; Comelli, D. Identification of pigments in different layers of illuminated manuscripts by X-ray fluorescence mapping and Raman spectroscopy. Microchem. J. 2016, 124, 775–784. [Google Scholar] [CrossRef]
  43. Colomban, P.; Lu, T.A.; Milande, V. Non-invasive on-site Raman study of blue-decorated early soft-paste porcelain: The use of arsenic-rich (European) cobalt ores-comparison with huafalang Chinese porcelains. Ceram. Int. 2018, 44, 9018–9026. [Google Scholar] [CrossRef]
  44. Colomban, P.; Simsek Franci, G.; Koleini, F. On-site Raman spectroscopic study of beads from the necropolis of Vohemar, Northern Madagascar (>13th C.). Heritage 2021, 4, 524–540. [Google Scholar] [CrossRef]
  45. Colomban, P.; Gironda, M.; Edwards, H.G.M.; Mesqui, V. The enamels of the first (soft-paste) European blue and white porcelains: Rouen, Saint-Cloud and Paris factories: Complementarity of Raman and X-ray fluorescence analyses with mobile instruments to identify the cobalt ore. J. Raman Spectrosc. 2021, 52, 2246–2261. [Google Scholar] [CrossRef]
  46. Analytical Methods Committee. Surface-enhanced Raman spectroscopy (SERS) in cultural heritage. Anal. Methods 2017, 9, 4338–4340. [Google Scholar] [CrossRef]
  47. Pozzi, F.; Leona, M. Surface-enhanced Raman spectroscopy in art and archaeology. Stud. Conserv. 2016, 61, 67–77. [Google Scholar] [CrossRef]
  48. Bruni, S.; Guglielmi, V.; Pozzi, F. Historical organic dyes: A surface-enhanced Raman scattering (SERS) spectral database on Ag Lee-Meisel colloids aggregated by NaClO4. J. Raman Spectrosc. 2011, 42, 1267–1281. [Google Scholar] [CrossRef]
  49. Leona, M. Microanalysis of organic pigments and glazes in polychrome works of art by surface-enhanced resonance Raman scattering. Proc. Natl. Acad. Sci. USA 2009, 106, 14757–14762. [Google Scholar] [CrossRef]
  50. Capitani, D.; Di Tullio, V.; Proietti, N. Nuclear magnetic resonance to characterize and monitor cultural heritage. Prog. Nucl. Magn. Reson. Spectrosc. 2012, 64, 29–69. [Google Scholar] [CrossRef]
  51. Blümich, B.; Casanova, F.; Perlo, J.; Presciutti, F.; Anselmi, C.; Doherty, B. Noninvasive testing of art and cultural heritage by mobile NMR. Acc. Chem. Res. 2010, 43, 761–770. [Google Scholar] [CrossRef]
  52. Blümich, B.; Casanova, F.; Appelt, S. NMR at low magnetic fields. Chem. Phys. Lett. 2009, 477, 231–240. [Google Scholar] [CrossRef]
  53. Madariaga, J.M. Analytical chemistry in the field of cultural heritage. Anal. Methods 2015, 7, 4848–4876. [Google Scholar] [CrossRef]
  54. Rehorn, C.; Blümich, B. Cultural heritage studies with mobile NMR. Anal. Chem. 2018, 57, 7304–7312. [Google Scholar] [CrossRef]
  55. Baias, M. Mobile NMR: An essential tool for protecting our cultural heritage. Magn. Reson. Chem. 2017, 55, 33–37. [Google Scholar] [CrossRef] [PubMed]
  56. Proietti, N.; Capitani, D.; Di Tullio, V. Applications of nuclear magnetic resonance sensors to cultural heritage. Sensors 2014, 14, 6977–6997. [Google Scholar] [CrossRef] [PubMed]
  57. Camaiti, M.; Bortolotti, V.; Fantazzini, P. Stone porosity, wettability changes and other features detected by MRI and NMR relaxometry: A more than 15-year study. Magn. Reson. Chem. 2015, 53, 34–47. [Google Scholar] [CrossRef]
  58. Brai, M.; Camaiti, M.; Casieri, C.; De Luca, F.; Fantazzini, P. Nuclear magnetic resonance for cultural heritage. Magn. Reson. Imaging 2007, 25, 461–465. [Google Scholar] [CrossRef]
  59. Zhang, Y.; Li, Y.; Liu, X.; Wang, Y.; Tang, H.; Qu, L.; Shang, Y.; Chen, W. Quantitative assessment of collagen degradation in archaeological leather by solid-state NMR. J. Cult. Herit. 2022, 58, 179–185. [Google Scholar] [CrossRef]
  60. Zhang, M.; Fan, J.; Liu, J.; Chen, Y.; Lu, Y.; Lei, Y.; Kaya, M.G.A.; Tang, K. A comprehensive evaluation of a historical leather armor from Yanghai cemetery, Turpan. Herit. Sci. 2024, 12, 162. [Google Scholar] [CrossRef]
  61. Moricca, C.; Stagno, V.; Hamza, N.M.; Favero, G.; Sadori, L.; Capuani, S. The high potential of micro-magnetic resonance imaging for the identification of archaeological reeds: The case study of Tutankhamun. J. Herit. Sci. 2023, 6, 7170–7180. [Google Scholar] [CrossRef]
  62. Morigi, M.; Casali, F.; Bettuzzi, M.; Brancaccio, R.; d’Errico, V. Application of X-ray computed tomography to cultural heritage diagnosis. Appl. Phys. A 2010, 100, 653–661. [Google Scholar] [CrossRef]
  63. Brittain, H.G. X-ray diffraction and x-ray fluorescence. J. Chem. Educ. 2006, 47, 177–226. [Google Scholar]
  64. Trojek, T.; Musílek, L.; Čechák, T. X-ray fluorescence analysis of cultural artefacts-applications to the Czech heritage. Radiat. Phys. Chem. 2014, 95, 381–384. [Google Scholar] [CrossRef]
  65. Galli, A.; Bonizzoni, L. Contribution of X-ray fluorescence techniques in cultural heritage materials characterization. Appl. Sci. 2022, 12, 6309. [Google Scholar] [CrossRef]
  66. Gómez-Morón, M.A.; Ortiz, P.; Martín-Ramírez, J.M.; Ortiz, R.; Castaing, J. A new insight into the vaults of the kings in the Alhambra (Granada, Spain) by combination of portable XRD and XRF. Microchem. J. 2016, 125, 260–265. [Google Scholar] [CrossRef]
  67. Daniel Martin-Ramos, J.; Zafra-Gómez, A.; Vílchez, J.L. Non-destructive pigment characterization in the painting little madonna of foligno by X-ray powder diffraction. Microchem. J. 2017, 134, 343–353. [Google Scholar] [CrossRef]
  68. Luo, W.; Si, Y.; Wang, H.; Qin, Y.; Huang, F.; Wang, C. Leather material found on a 6th B.C. Chinese bronze sword: A technical study. Spectrochim. Acta A Mol. Biomol. Spectrosc. 2011, 79, 1630–1633. [Google Scholar] [CrossRef]
  69. Franquelo, M.L.; Duran, A.; Perez-Rodriguez, J.L. Laboratory multi-technique study of Spanish decorated leather from the 12th to 14th centuries. Spectrochim. Acta A Mol. Biomol. Spectrosc. 2019, 218, 331–341. [Google Scholar] [CrossRef]
  70. Artem, S.M.; Nailyam, N.U.; Galina, V.P.; Maria, M.M.; Dmitriy, L.S.; Vitaly, V.P.; Dmitry, O.K.; Elena, I.D. Combination of total-reflection X-Ray fluorescence method and chemometric techniques for provenance study of archaeological ceramics. Molecules 2023, 28, 1099. [Google Scholar]
  71. Vázquez, C.; Albornoz, A.; Hajduk, A.; Elkin, D.; Custo, G.; Obrustky, A. Total reflection X-ray fluorescence and archaeometry: Application in the Argentinean cultural heritage. Spectrochim. Acta B At. Spectrosc. 2008, 63, 1415–1419. [Google Scholar] [CrossRef]
  72. Fernández-Ruiz, R.; García-Heras, M. Study of archaeological ceramics by total-reflection X-ray fluorescence spectrometry: Semi-quantitative approach. Spectrochim. Acta B At. Spectrosc. 2007, 62, 1123–1129. [Google Scholar] [CrossRef]
  73. Claudia Caliri, P.R.; Romano, F.P. Combined use of MA-XRF scanning and XRD for the non-invasive investigation of illuminated manuscript fragments. In Proceedings of the Technart 2017 International Conference, Bilbao, Spain, 2–6 May 2017. [Google Scholar]
  74. Liang, H. Advances in multispectral and hyperspectral imaging for archaeology and art conservation. Appl. Phys. A 2012, 106, 309–323. [Google Scholar] [CrossRef]
  75. Kavuru, M.; Rosina, E. IR Thermography for the Restoration of Colonial Architecture in India-Case study of the British Residency in Hyderabad, Telangana. J. Cult. Herit. 2021, 48, 24–28. [Google Scholar] [CrossRef]
  76. Hatır, M.E.; İnce, İ.; Bozkurt, F. Investigation of the effect of microclimatic environment in historical buildings via infrared thermography. J. Build. Eng. 2022, 57, 104916. [Google Scholar] [CrossRef]
  77. Tortora, M.; Sfarra, S.; Chiarini, M.; Daniele, V.; Taglieri, G.; Cerichelli, G. Non-destructive and micro-invasive testing techniques for characterizing materials, structures and restoration problems in mural paintings. Appl. Surf. Sci. 2016, 387, 971–985. [Google Scholar] [CrossRef]
  78. Mercuri, F.; Zammit, U.; Orazi, N.; Paoloni, S.; Marinelli, M.; Scudieri, F. Active infrared thermography applied to the investigation of art and historic artefacts. J. Therm. Anal. Calorim. 2011, 104, 475–485. [Google Scholar] [CrossRef]
  79. Laureti, S.; Sfarra, S.; Malekmohammadi, H.; Burrascano, P.; Hutchins, D.; Senni, L.; Silipigni, G.; Maldague, X.; Ricci, M. The use of pulse-compression thermography for detecting defects in paintings. NDT E Int. 2018, 98, 147–154. [Google Scholar] [CrossRef]
  80. Lanteri, L.; Agresti, G.; Pelosi, C. A new practical approach for 3D documentation in ultraviolet fluorescence and infrared reflectography of polychromatic sculptures as fundamental step in restoration. Heritage 2019, 2, 207–215. [Google Scholar] [CrossRef]
  81. Chu, Y.C.; Huang, Y.Y.; Ye, C.Y.; Chen, S.L. Automated wall moisture detection in heritage sites based on convolutional neural network (CNN) for infrared imagery. Appl. Sci. 2025, 15, 6495. [Google Scholar] [CrossRef]
  82. Wevers, M.; Nicolaï, B.; Verboven, P.; Swennen, R.; Roels, S.; Verstrynge, E.; Lomov, S.; Kerckhofs, G.; Van Meerbeek, B.; Mavridou, A.M.; et al. Applications of CT for non-destructive testing and materials characterization. In Industrial X-Ray Computed Tomography; Carmignato, S., Dewulf, W., Leach, R., Eds.; Springer: Cham, Switzerland, 2018; pp. 267–331. [Google Scholar]
  83. Vlassenbroeck, J.; Cnudde, V.; Masschaele, B.; Dierick, M.; Van Hoorebeke, L.; Jacobs, P. A comparative and critical study of X-ray CT and neutron CT as non-destructive material evaluation techniques. Geol. Soc. Lond. Spec. Publ. 2007, 271, 277–285. [Google Scholar] [CrossRef]
  84. Du Plessis, A.; Broeckhoven, C.; Guelpa, A.; Le Roux, S.G. Laboratory x-ray micro-computed tomography: A user guideline for biological samples. Gigascience 2017, 6, 27. [Google Scholar] [CrossRef]
  85. Laycock, S.D.; Bell, G.D.; Mortimore, D.B.; Greco, M.K.; Corps, N.; Finkle, I. Combining X-ray micro-CT technology and 3D printing for the digital preservation and study of a 19th Century Cantonese chess piece with intricate internal structure. J. Comput. Cult. Herit. 2013, 5, 1–7. [Google Scholar] [CrossRef]
  86. Fernandes, H.; Summa, J.; Daudre, J.; Rabe, U.; Fell, J.; Sfarra, S.; Gargiulo, G.; Herrmann, H.G. Characterization of ancient marquetry using different non-destructive testing techniques. Appl. Sci. 2021, 11, 7979. [Google Scholar] [CrossRef]
  87. Fuchs, T.O.J.; Hanke, R. 3D X-ray Tomography—Basics and Latest Developments. In Handbook of Advanced Non-Destructive Evaluation; Ida, N., Meyendorf, N., Eds.; Springer International Publishing: Cham, Switzerland, 2018. [Google Scholar]
  88. Stock, S.R. Microcomputed Tomography: Methodology and Applications; CRC Press: Boca Raton, FL, USA, 2019. [Google Scholar]
  89. Schwartz, L.F. Morphing the three faces of Mona: The decision-making steps Leonardo used to create his Mona Lisa. Comput. Graph. 1995, 19, 529–539. [Google Scholar] [CrossRef]
  90. Bettuzzi, M.; Morigi, M.P.; Brancaccio, R.; Peccenini, E.; Casali, F. A mobile computed tomography system for on-site cultural heritage analysis. In Proceedings of the IEEE International Conference on Environment and Electrical Engineering, Milan, Italy, 6–9 June 2017. [Google Scholar]
  91. Mannes, D.; Schmid, F.; Frey, J.; Schmidt-Ott, K.; Lehmann, E. Combined neutron and X-ray imaging for non-invasive investigations of cultural heritage objects. Phys. Procedia 2015, 69, 653–660. [Google Scholar] [CrossRef]
  92. Maher, M.A. X-RAY computed tomography of a late period falcon bronze coffin. Radiat. Phys. Chem. 2020, 166, 108475. [Google Scholar] [CrossRef]
  93. Machado, A.; Silva, A.; Campos, G.; Gomes, C.; Oliveira, D.; Lopes, R. Analysis of metallic archaeological artifacts by x-ray computed microtomography technique. Appl. Radiat. Isot. 2019, 151, 274–279. [Google Scholar] [CrossRef]
  94. Hoehse, M.; Paul, A.; Gornushkin, I.; Panne, U. Multivariate classification of pigments and inks using combined Raman spectroscopy and LIBS. Anal. Bioanal. Chem. 2012, 402, 1443–1450. [Google Scholar] [CrossRef]
  95. Pessanha, S.; Manso, M.; Carvalho, M.L. Application of spectroscopic techniques to the study of illuminated manuscripts: A survey. Spectrochim. Acta Part B 2012, 71, 54–61. [Google Scholar] [CrossRef]
  96. Janssens, K.; Van der Snickt, G.; Vanmeert, F.; Legrand, S.; Nuyts, G.; Alfeld, M.; Monico, L.; Anaf, W.; De Nolf, W.; Vermeulen, M.; et al. Non-invasive and non-destructive examination of artistic pigments, paints, and paintings by means of X-ray methods. Top. Curr. Chem. 2016, 374, 81. [Google Scholar] [CrossRef]
  97. Janssens, K.; De Nolf, W.; Van Der Snickt, G.; Vincze, L.; Vekemans, B.; Terzano, R.; Brenker, F.E. Recent trends in quantitative aspects of microscopic X-ray fluorescence analysis. TrAC Trends Anal. Chem. 2010, 29, 464–478. [Google Scholar] [CrossRef]
  98. Targowski, P.; Iwanicka, M. Optical coherence tomography: Its role in the non-invasive structural examination and conservation of cultural heritage objects—A review. Appl. Phys. A 2012, 106, 265–277. [Google Scholar] [CrossRef]
  99. Patrucco, G.; Gómez, A.; Adineh, A.; Rahrig, M.; Lerma, J.L. 3D data fusion for historical analyses of heritage buildings using thermal images: The Palacio de Colomina as a case study. Remote Sens. 2022, 14, 5699. [Google Scholar] [CrossRef]
  100. Raimundo, J.; Medina, S.; Mata, J.; Herrero-Tejedor, T.; Priego-de-los-Santos, E. Deep learning enhanced multisensor data fusion for building assessment using multispectral voxels and self-organizing maps. Heritage 2024, 7, 1043–1073. [Google Scholar] [CrossRef]
  101. Cera, V. Multisensor data fusion for culture heritage assets monitoring and preventive conservation. ISPRS Internat. Arch. Photogramm. Remote Sens. Spatial Inf. Sci. 2022, 46, 151–157. [Google Scholar] [CrossRef]
  102. Bilici Genc, B.; Bostanci, E.; Eskici, B.; Erten, H.; Caglar Eryurt, B.; Acici, K.; Ketenoglu, D.; Asuroglu, T. Development of a new non-destructive analysis method in cultural heritage with artificial intelligence. Electronics 2024, 13, 4039. [Google Scholar] [CrossRef]
  103. Marco, F.; Marina, K.; Massimiliano, P.; Arianna, T.; Alessio, D.B.; Stuart, J. Machine learning for cultural heritage: A survey. Pattern Recognit. Lett. 2020, 133, 102–108. [Google Scholar] [CrossRef]
  104. Tejedor, B.; Lucchi, E.; Bienvenido-Huertas, D.; Nardi, I. Non-destructive techniques (NDT) for the diagnosis of heritage buildings: Traditional procedures and futures perspectives. Energy Build. 2022, 263, 12029. [Google Scholar] [CrossRef]
Heritage 08 00381 g003
Figure 3. (a) The color analysis of a leather painting in the Alhambra’s Hall of the Kings in Granada, Spain [66], and (b) a summary of color identification.
Figure 3. (a) The color analysis of a leather painting in the Alhambra’s Hall of the Kings in Granada, Spain [66], and (b) a summary of color identification.
Heritage 08 00381 g003
Heritage 08 00381 g004
Figure 4. Applications of CT in (a) ancient marquetry revealing structural defects [86], (b) identification of Mona Lisa painting [89], and (c) exposure of hidden texture [93].
Figure 4. Applications of CT in (a) ancient marquetry revealing structural defects [86], (b) identification of Mona Lisa painting [89], and (c) exposure of hidden texture [93].
Heritage 08 00381 g004
Table 1. Comparative matrix of NDTs in CH.
Table 1. Comparative matrix of NDTs in CH.
CategoryAnalytical TargetSpatial
Resolution		PortabilityInvasivenessTypical CH Applications
Spectroscopic techniques
(FTIR, Raman, pNMR)		Organic binders,
Pigments,
Mental alloys		Point analysis:
10 μm (Raman) to 5 mm (pXRF)		Handheld (pXRF),
Benchtop (FTIR/Raman)		Non-invasivePigment ID, corrosion product analysis, organic material characterization
X-ray techniques
(XRD, XRF)		Crystalline phases,
Elemental composition		0.1–1 mm (lab XRD) to 3–10 mm (pXRF)Mobile systems availableNon-invasiveMineral identification, alloy composition mapping, authenticity verification
Spatially resolved X-ray
(μ-XRD, MA-XRF)		Elemental distribution10–100 μmLimited (benchtop)Non-invasiveHidden underdrawing mapping, pigment stratigraphy, degradation front visualization
Digital imaging
(HSI, MSI, IRT)		Surface features,
Sub-surface defects		30 μm (micro-HSI)-5 mm (IRT)Tripod-mounted systemsNon-invasiveMural deterioration mapping, hidden text recovery, moisture distribution monitoring
Volumetric imaging
(CT, μ-CT, Photogrammetry)		Internal structure,
3D morphology		0.5 μm (μ-CT) to 0.1 mm (CT)NoNon-destructiveBronze core casting analysis, mummy wrapping study, ceramic manufacturing technique reconstruction
Data fusion
(Heperspectral and CT, MSI and XRF)		Multi-scale propertiesN/APartialNon-invasiveCross-validated material diagnosis
AI-assisted diagnosisPattern recognition,
Predictive modeling		Pixel-level (segmentation)Cloud-based processingAlgorithmicAutomated crack detection, virtual inpainting, large-scale site monitoring
Table 2. A summary of the main diagnostic capabilities of FTIR.
Table 2. A summary of the main diagnostic capabilities of FTIR.
Material TypeDiagnostic Capability
Paper/CelluloseFiber sourcing, sizing agents
Textile (silk/wool/leather)Protein differentiation (fiber, keratin, collagen)
PaintingsMetal soap identification (Zn/Cu/Pb carboxylates)
Stone/CeramicsBinder degradation markers
Polymers3D artifact characterization
Table 3. Summary of AI learning paradigms in CH diagnostics.
Table 3. Summary of AI learning paradigms in CH diagnostics.
Learning ParadigmRepresentative AlgorithmsTypical ApplicationsKey Advantages
Supervised learningCNN, U-Net, Mask R-CNN, Transformer-basedCrack/defect detection,
material/region segmentation		High accuracy with labeled data
Unsupervised/
self-supervised learning		PCA, NMF, SOM, k-means, AutoencodersSpectrum clustering,
anomaly detection,
representation learning		Works without labels
Generative modelingGANs, Diffusion ModelsMural inpainting,
2D coin reconstruction,
3D virtual restoration		Realistic restorations,
immersive visualization
Semi-supervised
/transfer learning		—Leveraging limited,
domain-specific datasets for reconstruction or cross-domain detection		Reduces annotation cost,
improves generalization
Disclaimer/Publisher’s Note: The statements, opinions and data contained in all publications are solely those of the individual author(s) and contributor(s) and not of MDPI and/or the editor(s). MDPI and/or the editor(s) disclaim responsibility for any injury to people or property resulting from any ideas, methods, instructions or products referred to in the content.

Share and Cite

MDPI and ACS Style

Zhang, M.; Liu, S.; Shao, H.; Ba, Z.; Liu, J.; Albu Kaya, M.G.; Tang, K.; Han, G. Development Trend in Non-Destructive Techniques for Cultural Heritage: From Material Characterization to AI-Driven Diagnosis. Heritage 2025, 8, 381. https://doi.org/10.3390/heritage8090381

AMA Style

Zhang M, Liu S, Shao H, Ba Z, Liu J, Albu Kaya MG, Tang K, Han G. Development Trend in Non-Destructive Techniques for Cultural Heritage: From Material Characterization to AI-Driven Diagnosis. Heritage. 2025; 8(9):381. https://doi.org/10.3390/heritage8090381

Chicago/Turabian Style

Zhang, Mingrui, Suchi Liu, Haojian Shao, Zonghuan Ba, Jie Liu, Mǎdǎlina Georgiana Albu Kaya, Keyong Tang, and Guohe Han. 2025. "Development Trend in Non-Destructive Techniques for Cultural Heritage: From Material Characterization to AI-Driven Diagnosis" Heritage 8, no. 9: 381. https://doi.org/10.3390/heritage8090381

APA Style

Zhang, M., Liu, S., Shao, H., Ba, Z., Liu, J., Albu Kaya, M. G., Tang, K., & Han, G. (2025). Development Trend in Non-Destructive Techniques for Cultural Heritage: From Material Characterization to AI-Driven Diagnosis. Heritage, 8(9), 381. https://doi.org/10.3390/heritage8090381

Article Metrics

Back to TopTop