Information Technologies in the Research Activities of Museums: from Digital Catalogs to Artificial Intelligence Methods

Yumasheva Yuliya Yurijevna ORCID: 0000-0001-8353-5745 Doctor of History Deputy Director of "DIMI-CENTER" 105264, Russia, g. Moscow, bul. Izmailovskii, 43 Juliayu@yandex.ru Other publications by this author

Guk Dar'ya ORCID: 0000-0002-3380-9426 PhD in Philology Senior Researcher at the Department of Archaeology of Eastern Europe and Siberia of the State Hermitage Museum 190000, Russia, Saint Petersburg, nab. Palace, 34 hookk@hermitage.ru

With gratitude to colleagues,

standing at the origins of creation

domestic scanning equipment,

implementation of digitization projects

objects of historical and cultural heritage and

applications of electronic images in scientific research

The rapid development of information technology in the last three quarters of a century has had a huge impact on the transformation of all areas of human activity. Traditionally conservative organizations such as museums and archives, the beginning of the use of information technology in which dates back to the 1960s, did not stay away from these processes, but the use of IT received a special impetus in the mid-1970s – early 1980s. This is due, firstly, to the development and dissemination of database technology as the most successful model for creating electronic catalogs of objects of historical and cultural heritage, and, secondly, to the development of specialized equipment (scanners) that allows you to create electronic images that reproduce as fully as possible (visualizing in an electronic environment) the appearance of objects storage.

Leaving out of consideration the history of the use of database management systems (DBMS), which are the software basis for the formation of records of museum electronic catalogs, we will focus on studying and fixing the main stages of the history of the design of hardware and software complexes and the development of technologies for creating electronic images (copies) of objects of historical and cultural heritage and their introduction into the practice of accounting and stock, curatorial and research work of foreign and domestic museums ^[1], the analysis of the evolution of applied methods and technologies, the gradual expansion of the range of tasks of intramuseum activities in which they were used, the transformation of the forms of presentation of results, as well as the formation of an integral periodization of these processes. (Approx. authors: The authors leave out of consideration the issues of the use of electronic images in the exposition and exhibition work of museums, including Internet resources, with the exception of specialized projects closely related to high-quality electronic images).

To study the problems posed, a complex of types of materials and sources was formed, consisting of:

- monographs, articles, theses and other forms of publications of museum specialists, specialists in the field of information technology involved in solving specific tasks;

- descriptions and scientific reports on implemented projects aimed at the development and application of scanning methods, digital photographing, 3-dimensional digitization of museum monuments, and the use of electronic copies and developed technologies in scientific research (ATHENA, MINERVA (MInisterial NEtwoRk for Valorising Activities in digitalization), MINERAS PLUS, etc. ^{[2; 3; 4; 5]};

- materials of various international and national conferences (EVA, ADIT, etc. ^{[6; 7]}), round tables and seminars organized by specialized organizations (UNESCO, ICOM, CIDOC (ICOM Specialized Committee), etc.) ^[8];

- information systems of a reference and search nature and other digital resources presented online.

Analysis of the identified and studied literature (more than 2.5 thousand titles) published since the early 1970s (the first foreign - special issues of the UNESCO magazine "Museum" for 1971 and 1978 ^{[9; 10]} and articles by individual authors ^[11]; the first domestic publication – 1980 ^[12]) to date, has shown that there are no generalizing works in Russian and foreign historiography devoted to the study of the history of the introduction and application of information technologies in terms of creating electronic images of museum objects and their use in the intramuseum activities. (Approx. authors: the exception is the comprehensive research of D. Y. Guk ^{[13, 14]}; numerous textbooks and articles ^{[15; 16; 17; 18; 19; 20 and others]}, in which this topic is mentioned mainly in the context of using IT in working with visitors). The scientific literature, as a rule, describes the progress of specific projects, their development over a certain period of time, analyzes the activities of specialized conferences, seminars and supervising specialized organizations, which reveals the content of scientific research and discussions at the micro level, but does not offer methodological understanding and generalization of the results achieved on the scale of the museum community as a whole.

All of the above increases the relevance of considering the evolution and the proposal to periodize the introduction of electronic images of museum objects into intramuseum activities.

Before proceeding to the presentation of the results of the conducted historiographical research, it should be noted that from the mid-1960s to the present, the problem of creating and using high—quality images in museums has developed mainly within the framework of four directions: documenting the existence of museum objects (primarily archaeological monuments), electronic catalogues of museum collections ^[21] informatization and digitalization of technical and technological research (including restoration processes and research) and popularization of museum collections. Attention to the use of electronic images in each of these areas has been different in different periods, but in the last 20 years, due to the development of technology and technology (including the use of Data Science methods), documentation and scientific research have become a priority.

Further presentation will be conducted by periods, the criteria for which are the following aspects:

- objects that have been scanned (slides, radiographs, museum items directly);

- types of scanning and photo-fixing digital equipment (cameras, projection scanners, planetary contactless scanners for "direct" digitization of museum objects; 3D scanners, specialized multispectral cameras and scanners, etc.);

- image creation and image processing technologies and their evolution from traditional ("analog") methods of photofixation, multispectral photography, photogrammetry, radiography, spectromicroscopy, etc. to their digital (electronic) forms and varieties, as well as specialized software;

- electronic image formats (including formats approved as ISO standards);

- implemented projects within the framework of which the technology of digitization and scientific study of electronic images of objects of historical and cultural heritage was developed;

- the intended purpose of the created electronic images and their use in intramuseum (restoration, scientific preservation and research activities);

- user audience.

Separately, it should be emphasized that the proposed periodization, as well as the chronological boundaries of the selected periods, are not rigid and can shift in one direction or another for about 3-5 years, depending on historical circumstances and historiographical facts.

Apart from the topic of working with electronic images of museum objects is the problem of 3D models, virtual museums and reconstructions of movable and immovable cultural heritage (including archaeological). Taking into account the specifics of these areas, the authors will mention them in the text of the article only in cases of close proximity to the described topic.

Zero stage – mid-1970s – early 1980s. Statement of the problem of describing museum objects and supplementing electronic catalogues with images; scanning museum slide libraries with projection scanners

The development of automated text catalogues of museum collections has become one of the key topics in discussions held in the museum community since the early 1960s, when the first such catalog of the museum collection was created at the National Museum of American History (SELGAM) (1963). authors: the catalog has been reworked many times, and since the early 2000s it has been partially presented online: URL: https://americanhistory.si.edu/collections/about-online-collection ).

Almost simultaneously, the problem of creating museum catalogues was raised in the Soviet Union, where in 1965 the famous archaeologist Ya.A. Sher published an article in which he formulated approaches to the development of a database (in his terminology, a "cybernetic fund") ^[22] containing descriptions of archaeological sites.

However, the text catalogs had serious limitations, since the descriptions contained in them (entries in the database) did not make it possible to fully present information about each monument, as well as reflect the history of their existence in museums and study.

The problems of information incompleteness of museum catalogues have been actively discussed for a decade: in 1971, the UNESCO profile magazine "MUSEUM" prepared and published a special issue of "Museums and computers" (URL: https://unesdoc.unesco.org/ark:/48223/pf0000127370 ), in which the world's leading experts outlined approaches to creating the most complete descriptions of museum collections. By the end of the 1970s, as Ya.A. Sher noted, "one of the most pressing problems in the computer documentation of museum collections was the replacement of formalized descriptions with images with texts in natural language and the ability to work with them in an interactive mode ... ^{[23, p. 5]}".

One of the catalysts for the process of supplementing museum catalogues with images was a slide demonstration prepared by the staff of the State Hermitage Museum (Ya.A. Sher and A.O. Polyakov) and presented in May 1977 at a meeting of the CIDOC Committee of the ICOM General Conference held in GE ^[23]. The demonstration (even in the form of slides) made a splash, and although the project was never implemented in full, the report of Soviet museum archaeologists had a huge impact and probably launched the processes of developing specialized equipment and creating electronic images of museum monuments for electronic catalogs.

Simultaneously with the preparation of the demonstration, Ya.A. Sher worked on the creation of an original method for studying petroglyphs based on images. The approach proposed by him assumed the "recognition" and classification of elements of photo images translated into digital form. The study received full support from the Chairman of the Commission on the Use of Mathematical Methods and Computers in Historical Research at the Department of History of the USSR Academy of Sciences, corresponding member of the USSR Academy of Sciences I.D. Kovalchenko and specialists who worked in the Commission. The result of the works of Ya.A. Sher was an article published in a specialized collection, which for the first time in the history of Russian historical science described the method of using digitized images in the practice of archaeological research ^[24].

Another catalyst for these processes was the rich experience accumulated by many museums of the world by this time in creating images of museum objects on transparent media (negatives, transparencies, radiographs, etc.) and their use for exposition, publishing and restoration purposes (GIM, Alinari, Scala, etc. ^{[25; 26; 27]}). Such visual information has been accumulated for years in museum photo libraries, restoration departments or specialized structures (photo agencies and publishing houses), but it covered only a small part of museum collections. Photofixation was performed for recognized masterpieces or for solving specialized tasks. Total photography of museum collections was not carried out due to various reasons, primarily due to the complexity and capital intensity of the photo process. Even less often, specialized methods of radiography, spectromicrography, photogrammetry, etc. were used.

The development of technology for displaying static visual information in a digital environment (scanners) since the end of 1970 made it possible to initially raise the question of scanning the available information on transparent media. This task was solved with the help of projection scanners, which were actively used in high printing ^[28].

However, it soon became clear that the volumes of slide libraries and specialized restoration documentation are small, and photographs created at different times do not reflect the current state of objects, which, of course, could not satisfy either restorers, curators of museum collections, or researchers, especially since for the latter, the actual photograph has always been an "objective registrar of fact and the main research document ^[29]".

As a result of the combination of two independent factors that caused interest in the creation of electronic images, the issue of the need to develop hardware and software for "direct" color photofixation (documentation) of two-dimensional museum objects (paintings and graphics, archival documents, fabrics, objects of bonistics, numismatics, archaeological collections, etc.) was put on the agenda. in order to supplement them with electronic museum catalogues and use them in restoration and research practice.

The first stage is the development of contactless technology for direct photofixation of museum objects (mid–1980s – early 1990s).

One of the first to solve this problem was the VASARI Research Center for the Modernization of Equipment in the field of art and Technology at Birbeck College (Great Britain) ^{[30; 31]}. In the mid-1980s, the center's specialists developed a video camera (color scanner) for direct digital copying of paintings and a system for archiving and searching images. The kit also included a SUN workstation and a special BARCO monitor.

The scanner provided higher color accuracy and image quality (high resolution – 20 pixels per 1 mm, practically this is macro photography) than conventional photographic film, which allowed the use of created electronic copies in restoration and research work. The use of this equipment made it possible to create electronic images that gave a visual representation of the structure of the paint layer (layering and smears), pigments, their mixtures and grain distribution over an area of several square millimeters, and at the same time it was completely safe for the museum object, and also made it possible to organize in-line scanning, it was much more economical than the process of conventional photography. The resulting electronic image was a "passport" of a picturesque museum object, necessary for its unambiguous identification and accounting, study and determination of the state of preservation (physical condition) for restoration purposes.

The first use of the created scanner dates back to 1989-1992, when the equipment was installed in the London National Gallery ^[32]. The scanner made it possible to obtain images larger than 1 GB, which, in turn, required the creation of powerful image processing software (VIPS) ^[33] and (somewhat later) the development of a specialized server platform.

The quality and accuracy of the created electronic copies were so good that they began to be actively used in accounting and storage work (the images were integrated into the electronic catalog as a control image filling information gaps in the verbal description), for scientific analysis (including restoration and technological work on the study of textures and changes in time of media, writing materials of documents and the colorful layer of paintings ^{[34; 35]}), to ensure control over the safety of museum objects, to determine the need for their restoration and maintenance of this process, as well as for the publication of high-quality printed products. (Approx. authors: The possibility of analyzing the aging processes of paper and writing materials based on the study of photographic images was described earlier ^[36]) Electronic copies of paintings were initially stored on optical CD-R disks, and then in RAID systems.

The same images (in a reduced size) were used in the world's first museum electronic guide, released by Microsoft in 1993 on CD-ROM – "Microsoft Art Gallery on CD-ROM: The collection of the National Gallery, London». The publication was repeatedly reprinted and supplemented, and in 2002, based on it, the "Complete Illustrated Catalog of The National London Gallery on CD-ROM" (The National Gallery Complete Illustrated Catalog: Expanded (CD-ROM)) was released, which also used all the same electronic images of the paintings of the collection created a decade earlier.

Unfortunately, it was not possible to find publications of the results of scientific restoration and historical and art studies carried out with the help of this equipment. This is due to two reasons: the first is the unavailability of many professional journals online, which makes bibliographic search difficult; the second was that the main attention at that time was focused on studying the capabilities of the constructed equipment and using scanners for restoration purposes and for creating scientific catalogs, and not on historical and art research work.

Indirectly, this conclusion is confirmed by the beginning of the almost universal creation of electronic museum catalogs in museums in Western Europe and the USA. Not all of them were methodologically based on the use of high-quality electronic images as identifying elements, many still focused on the creation of standardized, thesaurus–based descriptions of museum objects - an approach that by the turn of the 1990s had an almost thirty-year history of discussion and was based on the use of database technology and various categorization schemes ^{[37; 38 etc.]}, or the compilation and use of dictionaries of iconographic analysis ^{[39; 40, etc.]}. But with the advent of the possibility of using images integrated into the database (at the turn of the 1980s and 1990s, the software implementation allowed specifying only the name of the image file stored separately in the text field) for unambiguous identification of objects and/or comprehensive representation of their iconographic content, the topic of scientific cataloging of museum collections received a new impetus.

Several European information search engines have become an example of the implementation of this approach, in particular, the French museum accounting system Joconde (URL: https://www.culture.gouv .fr/Thematiques/Musees/Les-musees-en-France/Les-collections-des-musees-de-France/Joconde-catalogue-collectif-des-collections-des-musees-de-France ) ^[41] and the electronic collection management system of the British Museum Merlin ^[42]. The development of both systems began in the mid-1970s, and they became available to the public in the early 2000s. (Approx. Authors: Since 2019, the Joconde system has been presented in the framework of the open platform of the cultural heritage of France – POP – URL: https://www.pop.culture.gouv .fr/search/list?base=%5B%22Collections%20des%20mus?es%20de%20France%20%28Joconde%29%22%5D; and public access to information about museum objects registered in the Merlin system was provided through the public access system to the Internet and kiosks of the British Museum COMPASS – URL: https://www.ssl.co.uk/ixbin/indexplus?record=PRO101 . Currently not functioning).

By the beginning of the 1990s, the first attempt to implement an intermuseum computer network was the European Museum Network project (EMN, 1989-1992), which united 8 museums in Europe and provided their visitors with multimedia resources at expositions, including texts, sound and images of 800 objects ^[43].

In Russia at the turn and in the first half of the 1990s, during the so-called "microcomputer revolution", the problems of digitizing works of art and museum collections and their scientific use and popularization were also given great importance. Thus, by the Resolution of the Council of Ministers of the USSR dated 10.01.1991 No. 28, the international consortium "Masterpieces of Art" was created with the participation of "Soviet organizations and foreign legal entities and citizens" ^[44]. The main objectives of the consortium were:

"creation of automated image banks and information networks, modern technical means of servicing visitors and storing museum valuables, assistance in the reconstruction of museums;

production and sale in the USSR and abroad of video films, videocassettes, optical discs, holograms, printed, souvenir and other products based on works of art and with artistic symbols....

conducting research and design work on the profile of the consortium's activities, professional development of personnel".

The State Russian Museum provided the museum and ideological component of the consortium's activities, and the technical implementation was provided by organizations and enterprises of the military-industrial complex involved in participation "in order to effectively use the conversion of the defense industry and joint entrepreneurship to strengthen the material and technical base and raise culture."

Unfortunately, the consortium focused its main attention on commercial activities, publishing, advertising and exhibition multimedia projects ^{[45; 46]}, popularizing the collections of the country's largest museums ^[47], and unique scanning equipment created by domestic specialists (including direct photo fixation - companies "Extel", "LOT", etc.) it was not widely used and was soon replaced by foreign products.At the same time, it is necessary to mention two projects implemented in Russia to create thematic digital collections using high-quality electronic images. The first is the computer data bank "Petroglyphs of Central and Central Asia" (1993), the authors of which were Kemerovo archaeologists under the leadership of Ya.A. Sher. The data bank was a full–fledged research project reflecting, on the one hand, the processes of documenting and studying archaeological sites, and on the other hand, it was a kind of result of long-term work begun in the mid-1960s to develop methods for cataloging and researching archaeological sites based on electronic images. Subsequently, numerous publications were published on the basis of this resource scientific works, a specialized website has been created and is still functioning ("Petroglyphs of Northern Eurasia" – URL: https://rockart-studies.ru/?p=3585 ) ^[48].

The second project was implemented in 1991-1992 at the State Tretyakov Gallery, where a text database with integrated images of paintings by I.K. Aivazovsky was formed on the basis of digitization of catalog information and slides. The project was presented in 1992 at the Prado Museum in Spain at an exhibition dedicated to the 500th anniversary of Columbus' discovery of America. The creation of the database was considered by its creators as the first (main) stage in the implementation of a larger program for the formation of thematic knowledge bases dedicated to the work of individual artists, which the authors of the project intended to create and fill. Unfortunately, the program did not receive the support of the professional community and was only partially implemented, and by the mid-2000s, despite attempts to rethink it at a new stage and resuscitation ^[49], it was completely closed.

The second stage is the development of specialized scanning equipment designed to work with UV and IR radiation and 3 D models (1994-1999).

The next stage in the development of electronic image creation technologies is also associated with the activities of English specialists, modification of the VASARI scanner and its software ^[50]. New versions of the scanner allowed to work with ultraviolet and infrared radiation. In fact, the developed equipment was supposed to digitalize and automate the technologies of multispectral photography and radiography, which are known to museum specialists since the first half to the middle of the XX century. ^{[51; 52; 53; 54]} and have proven themselves well as methods of studying painting (including fresco); archival documentation – palimpsests, fading texts, photographs; archaeological artifacts, and were also used in paleographic and filigree studies. The created scanner was intended for use not only in the National London Gallery, but also in the Uffizi Gallery in Florence (the European MUltimedia Special Actions project (MUSA–ESPRIT), 1994-1999) ^[55]. (Approx. authors: it should be noted that at the same time, other projects were underway to develop multispectral digital cameras ^[56]).

The purpose of the new hardware and software complex was to provide digital support for the preservation, conservation and restoration of museum objects ^[57], i.e. the equipment did not change anything in the methodology of automated technical and technological processes and procedures, it only accelerated them and provided opportunities for the development of new digital services that could potentially lead to the emergence of new research methods Thus, the complex allowed collecting the results of the diagnostic examination of paintings in order to ensure their "comparative reading" and monitoring their changes over time. As part of the project, a color certification software system was developed ^[58], which made it possible to compare the color of digital images with the original work of art ^[59]. The resulting digital images were also integrated into the Uffizi database and became the visual basis of the gallery's record-keeping and research work ^[60].

Another project carried out at the VASARI Center was the ACOHIR project, which developed tools for obtaining images of 3D objects in museums and high-resolution and accurate color rendering software for viewing them. Unfortunately, this 1993 project, which was significantly ahead of its time, is practically unknown to specialists, although the equipment (a turntable with a digital camera) and software were used in Southampton Museums and in the Louvre (The ACOHIR – URL: https://www.southampton.ac.uk /~km2/projs/acohir/ ^[61]).

Almost simultaneously with the work of English specialists, in 1993, a parallel European research project was launched in continental Europe, the main purpose of which was to create a bank of high-quality electronic images of museum objects intended for conducting art history and historical research and the formation of thematic databases (Network of Art Research Computer Image SystemS in Europe – NARCISSE). The project, based on the iconographic classification of museum objects developed by the Dutch specialist G. van Waal ^[62], used Thomson-Broadcast projection scanners, which allowed working with radiographs, with photography performed using UV and IR filters, and creating huge files intended for scientific experiments ^[63]. The results obtained in the NARCISSE project made it possible to move on to solving more complex tasks at the next stage of the development of technologies for working with electronic images of museum objects (within the framework of the European Research Open System, EROS project).

Both projects – MUSA-ESPRIT and NARCISSE – were aimed at translating traditional technologies (rengtenography, spectromicroscopy, multispectral photography, etc.) into a digital environment, improving research methods for museum objects (primarily paintings), developing methods for dating and detecting fakes, clarifying the state of preservation of museum objects, forming technical specifications for conservation and restoration and monitoring of restoration processes (including for the formation of appropriate knowledge bases on restoration problems and techniques), as well as for the use of the obtained images in high-quality printing (MARC project ^[64]).

A significant addition to these projects was the commercial DeBabelizer software, developed in 1992 by Equilibrium Technologies, Inc. The program allowed processing more than 50 types of raster graphics files and forming so-called "averaged palettes" based on arrays of the same type of files. This technical solution, on the one hand, made it possible to significantly save space for storing large amounts of graphic information, and on the other hand, it was the basis of the research method of establishing the authorship and authenticity of paintings, since it allowed creating individual color "author palettes" inherent in specific artists.

Since the mid-1990s, publications began to appear in specialized periodicals, in which the results of the use of the created scanning systems and methods of analyzing electronic images of paintings were presented. Among such studies, it is worth mentioning, for example, the works of specialists from the National London Gallery devoted to the study of the "Garvagh Madonna" by Raphael ^[65] and the famous portrait of the Arnolfini couple by Van Eyck ^[66], as well as the analysis of the influence of Leonardo da Vinci's creative method on the paintings of his pupil Giampietrino Boltraffio ^[67]..

It cannot be said that from the point of view of setting research tasks, these were absolutely original works, however, the use of computer technologies (specialized scanning equipment and image processing software) accelerated research processes and provided new additional opportunities for extracting and analyzing information from museum objects, which played a significant role in the development and dissemination of these technological solutions.

At the same time (1996), the publication of the world's first fully digitally prepared colorimetric scientific catalog of works of art – the collection of Flemish Baroque paintings of the Munich Pinakothek, created as part of the MARC project ^{[68; 69]}.

The experience of developing scanning systems and software for obtaining high-quality electronic images of museum objects, methods of their automated analysis and publication only in less than a decade (from 1994 to 2002) were summarized and presented to the public in four professional reviews, the authors of which were specialists who worked in various European projects ^{[70; 71; 72; 73; 74]}. In these articles, the authors outlined the prospects for the development of technologies for creating and using electronic images in scientific research, considering them (images) as full-fledged historical sources with specific quantitative characteristics suitable for applying statistical analysis methods to them, which anticipated the use of Data Science methods in art history and art history.

Considering the history of the development and application of electronic images in the mid – second half of the 1990s, it is necessary to note several more trends that emerged during this period.

Thus, at that time, interest in the "digital adaptation" of another technology known since the middle of the XIX century ^[75] – photogrammetry ^{[76; 77]}, which became one of the leading methods of research of archaeological collections, significantly increased.

Projects to create catalogues-databases with integrated images of museum objects, and, above all, paintings, continued and developed. A certain innovation of this period was a purely historical thematic project of the Institute of Everyday Realities of the Middle Ages and Early Modern Times of the Austrian Academy of Sciences, which assumed the formation of a database based on segmentation (highlighting details) on electronic images of household paintings and descriptions of each selected element in order to further process and "establish a connection and hierarchy not only of formal attributes of canvases, but also and their plots." The formulation of the research task did not differ in novelty – such work was carried out earlier, but the use of the KLEIO DBMS and the IBM IMAGE ASSISTANT software product for processing artistic images for the purposes of the project, which made it possible to link an image and text information within one REAL information system, were new ^{[78; 79; 80]}.

Projects related to the solution of telecommunications problems and the integration into a single network of museums located in various European countries and the exchange of information about museum objects, including their electronic images, have also found their development. We are talking about the projects RAMA (Remote Access to Museum Archives), VAN EYCK (Visual Arts Network for the Exchange of Cultural Knowledge) and Viseum.

The purpose of the first project was to create a unified system of remote access to the scientific reference apparatus and electronic copies (high-quality images) of archival documents stored in the museum archives of Berlin, Paris, Florence, Oxford and Madrid ^{[81; 82]}.

The second project was devoted to the formation of a computer network for the exchange of information about artistic values and was developed in the interests, first of all, of the professional scientific community and museum staff ^[83].

The main task of the third – Viseum ^[84] – was the development of advanced Internet applications for museums. In particular, a high-speed network (ATM) was created between Vancouver, Berlin, Paris and London. This made it possible to quickly create interactive images with high resolution and accurate color images of works of art. Methods have also been developed to search for data from various sites by common search attributes, and authorization methods have been introduced to control access, allowing subscribers to provide valuable data. One of the results of this project was the creation of a tiled pyramid TIFF format for presenting high-quality images on the Internet ^[85].

It is not difficult to guess that together with the previously implemented EMN project, these projects were the forerunners of the Europeana platform, implemented almost 15 years later.

An unexpected purpose of the banks and databases of electronic images of museum objects created since the early 1990s was the fight against the smuggling of art valuables. In order to prevent illegal trafficking of museum objects, Interpol and the International Organization for Standardization (ISO) developed and adopted the ISO/IEC-10918 standard, which became the standard for museum databases with electronic images integrated into the database structure (this technical solution was an innovation of the period under review). The standard included uniform rules (a set of fields, requirements for filling them in) for the minimum description of a museum object and the format of electronic images Jpeg-10918 ^[86]. The electronic image was indexed by attaching a standardized short description directly inside the image file. On the basis of ISO-10918, a single European network of computer images "Museum On Line" (MOL) was created, as well as a package of service programs, in the development of which specialists from universities, research centers and IT companies from France, Germany, Great Britain, the USA and Japan participated ^[87].

The ISO-10918 standard has also found its application in Russia, where, on its basis, the State Historical Museum carried out work on the creation of a Security videographic data Bank of museum storage items ^{[88; 89]}, – in a sense, the forerunner of the State Catalog of the Museum Fund, – and the Russian node of registration of computer images ISO RU1095 was formed ^{[90; 91; 92]}, designed for e-commerce tasks. The creation of electronic images for these purposes was carried out either on projection scanners (i.e. from transparent media, or by the video input method ^[93]). (Approx. Authors: A similar project was carried out within the European Union in 1996-1998: MENHIR – Multimedia European network of high quality image registration. URL: https://cordis.europa.eu/project/id/24378/it).

However, as before, the most important areas of the use of electronic images in museum activities remained exposition and exhibition work, the creation of multimedia publications, which experienced in the second half of the 1990s. the so–called "multimedia boom" ^{[94; 95]}, and the development of museum sites is a new trend that appeared in the late 1990s.

The largest Russian museums were at the forefront of site building in Russia: the State Hermitage Museum, the State Museum-Reserve "Moscow Kremlin", the State Russian Museum, etc. ^[96], whose websites, in addition to mandatory information elements, sometimes hosted innovative resources. For example, the first version of the GE website, created in 1997-1998, hosted coloristic and compositional image search services developed by IBM on the basis of electronic copies of paintings from the museum's collection (QBIC software – image content query ^[97]), which was an unconditional step forward in terms of studying museum objects storage with the help of computer technology, and in terms of popularization of museum collections.

The third stage – the turn of the third millennium – the 2000s. Changing methodological approaches to electronic images; widespread use of "direct" scanning scanners, web technologies, active museum site building, increasing the scale of digitization.

The late 1990s – early 2000s were marked by serious changes in the study of electronic images of objects of historical and cultural heritage. During a relatively short period in the professional community, there was a methodological rethinking of the electronic image as a set of individual pixels (in fact, a set of data), the color of which is determined in a specific color space (RGB, CMYK, HSV, etc.). Thanks to this "quantitative" specificity of the histogram of electronic raster images (graphical representation of the distribution of colors, textures, compositions, etc.) began to be considered as one of the ways of representing (modeling) image information. In addition, it became obvious that histograms can be calculated as an approximation of the color distribution, which gives a simple and compact representation of the coloristic, textural, geometric, etc. content of a work of art. (Approx. authors: the texture analysis tool was built on the basis of Pass and Zabikh histograms ^[98] and Haralik coincidence matrices).

This methodological approach based on statistics had two serious consequences.

The first was to understand the direct dependence of the correctness of the conclusions made on the basis of the analysis of electronic images on their quality (image size in pixels, color accuracy (the number of displayed shades of colors), etc. In both cases, the larger (size, number of colors), the better), and set the task of developing tools (scanners, cameras and other equipment) to obtain the most adequate originals (i.e. high–quality) images, as well as monitoring the processes of creating such images.

The second consequence is the possibility of using a wide range of Data Science methods in the study of electronic images of museum objects, which in turn led to an exponential growth of scientific developments and publications on the subject of creating image search systems (for example, the PICASSO system, which used a pyramidal image format, segmentation and indexing of colors as a basis for the formation of queries ^[99] , the definition of individual author's "palettes" characterized by a certain set of colors (not only on the basis of "averaging" the number of color shades, as it was implemented in DeBablizer), confirmation on their basis of authorship, authenticity, clarification of dates, etc. Thus, the first approved methods of analysis included:

- cluster analysis (K-means algorithm ^{[100; 101]}), mode search algorithm (Mean-Shift) ^[102];

- Blobworld image segmentation method ^{[103; 104]};

- image decomposition method (including discrete cosine transform, wavelet transform, Gabor filters ^{[105; 106; 107; 108]} ( etc.).

During this period, the active development of a new branch of historiography, which until the turn of the century was on the periphery of scientific search, became especially noticeable — the study of electronic images as an art and purely historical (material, pictorial) source.

At the same time, other organizational and technical changes took place during this period, which gave the decade decisive importance in the scientific use of electronic images. These changes include:

- improvement and wide distribution of industrial scanning equipment, including "direct" 2D scanning scanners (developer firms: Cruse, Microbox, Metis, Zeutschel, etc.) and 3D scanners (developer firm Breuckmann, etc.);

- design and creation of multispectral photography cameras, digital radiography, specialized 3D equipment, gigapixel photos (including panoramic ones – Gigapxl™ camera – Gigapxl Project – URL: http://www.gigapxl.org /) and their application in research practice (CRISATEL project ^{[109; 110]}, Lumiere company; ^{[111; 112]});

- development of specialized methods for monitoring scanning processes, test objects, and the quality of equipment settings (primarily in projects for digitizing archival documents and book monuments ^{[113; 114; 115; 116; 117; 118; 119; 120 etc.]}), as well as automated control programs to guarantee the quality of the created electronic copies ^{[121; 122]};

- development of methods and specialized formats for the presentation of electronic graphic information (pyramid formats based on wavelet transformations, for example, JPEG-2000 (see, ISO/IEC 15444 series of standards)), focused on the publication of high-quality images on the Internet;

- active development of interstate (EU – "Electronic Europe", 1999; UNESCO – "Information for All", 2000) and state ("Electronic Russia", 2002) programs in the field of information and communication technologies, the activities of public associations focused mainly on increasing the scale of digitization of historical and cultural objects heritage (the Lund Principles of 2001 and the Parma Charter of 2003), the use of web technologies and the construction of websites of museums and/or international museum associations that presented digitized images of sets of objects from museum collections (projects MINERVA (MInisterial NEtwoRk for Valorising Activities in digitisation), MINERVA PLUS and BRICKS), virtual museums ^[123;124], etc.

Unfortunately, the historical and artistic scientific component of these international programs has practically disappeared, and research has been carried out by individual museums and research centers on an initiative basis without any coordination of scientific activities, but with an active professional discussion of the problems being solved at the sites of scientific conferences and specialized periodicals, mainly in the technical sphere.

One of the few exceptions to this rule was the EROS project (European Research Open System – URL: http://www.numerique.culture .fr/pub-fr/document.html?base=dcollection&id=FR-DC-M6001_003), which was supervised by French specialists from the Center for Research and Restoration of Museums in France (C2RMF). The aim of the project is to create a multilingual database designed to manage all types of digital information (data, documentation, images) created in relation to objects from museum collections. The database has accumulated all analytical accounting and restoration data, all scientific reports and publications, digital quantitative analysis data, spectra, graphs, chemical formulas, all images of museum objects (including UV, IR, directional light photography, electron microscopy images, panoramic photography, X-rays, all images collected during the period of existence of objects in the museum, etc.), the latest technological developments in the digital sphere and the Internet, including automatic recognition of images or text content on objects. In fact, the database model implemented in the EROS project was a full-fledged knowledge base that would accompany any museum collection and provide an opportunity to work with a wide range of documentation and metadata, both already formed and continuing to be formed and characterizing each item individually and the collection as a whole ^{[125; 126]}. At the same time, document and metadata complexes could easily be transformed into data sets (data wrangling), which gave the database a completely new quality – the ability to work with the information contained in it using Data Science methods. Currently, it is still possible to find individual elements of the mentioned system on the Internet (the "pilot" collection of ancient Greek archaeology), but the project as a whole is not presented on the Internet.

Such an ideology of automated collection management was the basis for the work carried out at the State Historical Museum in the mid-2000s. Computerization programs of accounting and stock work, in which a prominent place was occupied by work with digitized historical photographs and modern electronic images of museum objects ^[127]. Unfortunately, in 2010, the work on this program was curtailed.

At the same time, despite the reorientation of large projects to web technologies in the professional circles of museum workers in the 2000s. the interest and relevance of using electronic images as a basis for automated historical and art history analysis, virtual restoration and reconstruction, image search (coloristic, compositional, semantic analogies), research of artistic practice and creative heritage of individual authors, authentication of museum objects, authentication, security, etc. has significantly increased ^[128]. This interest was supported by professional periodicals (ACM Journal on Computing and Cultural Heritage, Journal of Mathematics and the Arts), special issues of general journals ^{[129; 130; 131]}, regular conferences (EVA conferences, Museums & the Web, Eurographics Computational Aesthetics, CHArt, etc.) or special projects ^{[132; 133]}, the activities of working groups (DELOS/NSF, Digital Painting Analysis Group (Group. 2009. Digital painting analysis – URL: http://www.digitalpaintinganalysis.org .), etc.) and Technical Committees (IAPR TC19 Computer Vision for Cultural Heritage Applications – URL: https://www.cvl.iis.u-tokyo.ac.jp/IAPR-TC19 /).

During this period, projects of the 1990s related to the creation of high-quality electronic images, their presentation now in online resources intended for specialists, as well as the improvement of methods for studying museum objects based on electronic images and digitalized technologies of multispectral analysis, radiography, the development of knowledge bases based on on electronic images, Data Science methods, including artificial intelligence.

So, in particular:

- specialists of the National London Gallery worked on the development of software (using computer image modeling methods) for digitizing and studying radiographs ^[134]; they worked out mechanisms for layer-by-layer study of painting from the canvas base to varnish based on computer analysis of electronic copies created in the IR spectrum, which allowed to identify the evolution of the artist's idea ^[135]; improved scanning equipment and carried out with its help studies of changes in paintings by the authors ^[136]; proposed methods for monitoring the safety and restoration processes of museum items ^[137];

- French specialists of the Center for the Restoration of Museum Collections actively engaged in solving the problem of presenting high-quality images on the Internet ^[138]; developed and used digital multispectral cameras for layer-by-layer analysis of famous paintings (in particular, Leonardo da Vinci's "Gioconda" ^{[139; 140]});

- the curators of the Netherlands Van Gogh Museum, together with programmers from the University of Maastricht, used artificial intelligence methods (measurement of additional colors and the Gabor filter) to analyze a data set formed from 617 electronic copies of Van Gogh's paintings to confirm the hypothesis of a change in the artist's artistic technique, thereby proving the possibility of using quantitative (statistical) methods to analyze painting, as well as confirming the objectivity of the results obtained. ^{[141; 142]};

- employees of UNESCO and major scientific organizations developed mechanisms for creating and managing knowledge bases of museum collections ^[143], as well as image search by context and image metadata ^[144], etc.

In Russia, such developments were carried out mainly in restoration centers and were carried out on a much smaller scale due to the lack of appropriate equipment, insufficient funding for the museum sector and priority attention to the introduction of information technology in working with visitors.

The question of the practical use of electronic images in intramuseum activities was actualized in the mid-2000s after the tragic events at the State Hermitage Museum in connection with the search for ways of security marking and unambiguous identification of museum items. In 2007-2008, at the State Tretyakov Gallery, under the leadership of Deputy Director S. A. Khrustalev, the possibilities of micro and macro photography were studied, the results of the well-known European Fingerprint project were analyzed, in which 2D spectral images were combined with data obtained by 3D scanning of the surface of museum objects to form unique identifying data – the so-called "fingerprints" of objects that are not can be reproduced on fakes (URL: https://www.southampton.ac.uk /~km2/projs/fingartprint/index.html ). Unfortunately, the proposals of the GTG did not receive the support of the Ministry of Culture, and the work was completed.

Another project carried out in the late 2010s by the Pushkin State Museum of Fine Arts together with the Epos Group - high–quality digitization of museum collections for their subsequent presentation on the Internet – was more successful. The ideological basis of the project was an idea that had been actively developed for a decade in foreign museums, but for many museum specialists in Russia it needed at least verbalization, at most promotion and protection. (As one of the leaders of this project V.V. Definenov wrote: "The problem of using digital images is solved by creating thematic databases based on Internet technologies, which will not only expand the attributive scientific base, but also increase the total volume of available collections" ^[145]).

The result of the project at the State Museum of Fine Arts were seven thematic "satellite sites" of the museum's main resource, on which digitized collections were placed ("Electronic Collections" URL: https://www.pushkinmuseum.art/media/e-coll/index.php?lang=ru ), and a patented digital photography technology that did not have a harmful effect on the originals and made it possible to create high-precision digital archives of paintings, graphics, manuscripts, icons and objects of decorative and applied art.

Summarizing the consideration of the history of the use and study of electronic images in the 2000s, it is worth paying attention to the obvious heterogeneity of the events of this decade, expressed in a sharp expansion and change of paradigms, subjects and objects of research (not only restoration and art history, but also general history; not only "flat" paintings, graphics, etc. but also objects of decorative and applied art) and in attracting specialists of various professions (not only restorers and/or art historians, but also historians, computer scientists, specialists in the field of optical systems and mathematical research methods), and in redistributing the tasks to be solved between software and the equipment being created. The last thesis needs a comment: for example, if in the early and mid-2000s the question of the possibility of software processing of two-dimensional electronic images in order to form 3D models based on them was discussed ^[146], then by the end of the 2000s. this task was practically eliminated due to the appearance on the market of specialized scanning equipment and software complexes for it, which created 3D models with the reproduction of not only geometric volumes, but also textures and coloristic components.

The fourth stage – 2010s – present. New possibilities of scanning and photo-fixing equipment and software, the use of Data Science methods for the analysis of electronic images.

The features of the historiography of the creation and study of electronic images of museum objects, outlined in the third period, were developed in the 2010s.

Among the main trends of the fourth period , it should be noted:

- expansion of the research topics of electronic images: from restoration practices, identification of authorship/forgery (authentication), features of the artistic method ^[147], clarification of dating ^[148], to virtual archaeology ^[149] – modeling/reconstructions of individual objects and the historical environment of their existence (see, for example, conference materials: Computer Applications and Quantitative Methods in Archaeology – URL: https://caa-international.org /; Virtual Archeology (State Hermitage Museum) – URL: http://www.virtualarchaeology.ru/index.php );

- improvement of scanning equipment, software, the emergence and/or adaptation of specialized methods of image fixation (structured light scanning (SLS), computed tomography ^{[150; 151; 152; 153]}, new methods of photogrammetry (SfM) ^[154], portable spectrometers ^[155], etc.) and their combinations, gigapixel photography), digital retgenography, platforms for the presentation of electronic images (including by means of VR/AR);

- active use of Data Science methods (visualization, computer modeling, computer vision ^[156], artificial intelligence, big data technologies, etc.) to work with images of museum objects and their collections ^{[157; 158; 159]};

- shifting the emphasis in the study of museum objects from the textual (descriptive) component to the image/images (including the prevalence and popularity of 3D models as a means of documenting collections ^{[160; 161, etc.]} and their popularization (for example: The Uffizi Digitization Project – URL: http://www.digitalsculpture.org/florence/main/main_collection – 300 3D models of sculptures from the collection of the Uffizi Gallery; The Virtual Hampson museum – URL: http://hampson . cast.uark.edu – 3D models of archaeological artifacts; Smithsonian X 3D – URL: http://www.3d.si.edu – 3D models of the Smithsonian Museum collection), including 3D printing ^[162]), the emergence in the IT industry and the reinterpretation in the museum sphere of the concept of "digital twins" (Digital Twin) museum objects and spaces (Approx. 1) as surrogates imitating originals, etc.);

- and, finally, the question of ways and methods of ensuring the preservation and further use of digital heritage, including born-digital works of art and museum digital resources, which are based on electronic images of museum objects ^{[163, 164]}, the number of which is increasing day by day, has risen in full growth.

Not being able to consider in detail all these areas, we will limit ourselves to mentioning several materials and regulatory documents adopted by various museum institutions ^{[165; 166]} and focus on several landmark projects related mainly to new technologies for obtaining high-quality images, specialized software and the study of electronic images using DS.

One of the significant phenomena (trends) in terms of creating high-quality images was the development and appearance at the turn of the 2010s of the technology of gigapixel panoramic photography, the pioneers of which in museums are Google/NASA /CMU (Carnegie Mellon University). To date, there are several hardware and software solutions (GigaPan, GigaPixel, GIGAmacro, Photosynth, etc.) used to create digital photos and panoramas of super-high resolution. The formation of these images is still extremely time-consuming and requires expert knowledge, which prevents the widespread use of gigapixel images for research purposes. The most impressive exception to this rule are projects documenting archaeological sites, in particular petroglyphs, which began to be implemented in the late 2000s - early 2010s: for example, gigapixel photo fixation of petroglyphs of Saudi Arabia (GigaPan Imaging – Arabian Rock Art Heritage – URL: http://saudi-archaeology.com/imaging-techniques/gigapan-imaging /; http://www.gigapan.com/gigapans/46668 ), or creating a gigapanoram the image of the plate on the Devil's Nose, made by the archeologists of the Hermitage (URL: http://www.archaeoglobus.sfu-kras.ru/monument/petroglify-na-besovom-nosu /; http://virtualarchaeology.ru/full/index.php?id=225 ).

Nevertheless, since the beginning of the 2010s, examples of gigapixel images have been widely presented online in educational projects:

- Zoom Views — Google Art & Culture (URL: https://artsandculture.google.com/project/gigapixels ) – 1859 gigapixel images of works of fine art from around the world;

- in the collection of the Italian company Halta Definizione (URL: https://www.haltadefinizione.com/image-bank /), specializing in high-quality gigapiskel digitization combined with 3D scanning of museum objects and archival collections for restoration, conservation and study purposes;

- in a virtual spherical gigapixel panorama of the Sistine Chapel of the Vatican – URL: https://www.vatican.va/various/cappelle/sistina_vr/index.html;

- virtual panorama "Opening of the Kremlin – virtual tour of the residence of the President of Russia" – URL: http://tours.kremlin.ru/а etc .

The mid-2010s were marked by further development and improvement of the technology of gigapixel photography and scanning. So, in 2016, employees of the Swiss Federal Institute of Technology in Lausanne and ARTMYN presented 5D scanning technology, which allows "stitching" 100 high-precision scans of an object into one model (URL: https://artmyn.com/products/interactive-5d-image /). ARTMYN scans show their textures and materials, allowing you to "touch with your eyes" the more than 1.2 billion pixels that make up the objects. Digital copies (collection of 200 electronic copies of museum objects – URL: https://www.artmyn.com/explore ) can be studied and analyzed on a screen with lighting positioned at any angle, i.e. almost as if the originals were in the hands of the researcher. To present these images, in collaboration with fine art restorers, ARTMYN has developed a new Digital Condition Reports (DCR) standard, where all the features of an artwork can be described and commented on by experts and specialists in a very detailed, technological, interactive mode. At the same time, the IR and UV parts of the spectrum can also be used for this work (imitation of multispectral photography – URL: https://www.artmyn.com/products/multispectral-analysis /). In the course of this work, the program generates a protocol – a digital report, which is a unified and transparent way of analyzing works of art presented on the screen. The created copies can be used for security or insurance purposes, since the scanning process also generates a digital image of the original in the visible and ultraviolet spectrum. Thus, the very unique "portrait" of the object is created, including topographic, reflective and colorimetric properties. The technology reacted to the slightest changes in the state of the original, invisible to the naked eye, and, therefore, this copy becomes a guarantee of protection of the original from substitution.

No less impressive were the hardware solutions and application software, which in the early 2010s were created by such well-known manufacturers of scanning equipment as Cruse, Metis and Microbox.

The first of them developed a technology for direct digitization of paintings without extracting the original from a glazed frame. The second (Metis), developing technologies for digitizing documents of the Secret Archive and book monuments of the Apostolic Library of the Vatican, offered a comprehensive format of high-quality electronic images (METIS "MDC") and a number of software applications for its processing (URL: https://www.metis-group.com/metis-software ), allowing to obtain from a single digitized file any electronic images necessary for the researcher – multispectral, relief, 3D, etc., which ensures the safety of the original. The third (Microbox) has launched a line of reprosystems that make it possible to obtain images of hidden information – filigrees, palimpsests, spolveros, etc. (URL: https://book2net.net/ru/produkte/репрографические-системы /), using specialized equipment for this purpose.

In the winter of 2022, English specialists from the Bodleian Library of Oxford, together with engineers from the Spanish company Factum Foundation, announced the launch of a unique project ARCHiOx (Analysis and Recording of Cultural Heritage in Oxford – Analysis and Accounting of the cultural heritage of Oxford – URL: https://oxford.shorthandstories.com/digital-archiox/index.html ?fbclid=IwAR2LM19j6iFh1NUgEBddBmU0oZotufAEEs8G0vn2FzF97_dFd2c-TUUWGBS), within which a unique scanning equipment was designed and created to solve the problems of complex study of museum, archival and library objects (URL: http://blogs .bodleian.ox.ac.uk/theconveyor/archiox-research-and-development-in-imaging/?fbclid=IwAR0wT4zxCOmdvHT5LORdT-Zhyp-n_nyVR5wW0wuzzXzXDLEAqWt91UtV4CA). Describing the features of the equipment and the methods being developed, one of his project managers, John Barrett, notes that "two modern cameras used in ARCHiOx use different principles for recording volume and are specially designed to capture the texture of a low-relief surface. This approach is best suited for digitizing predominantly flat, but texture-rich originals from the Bodleian collections." Such shooting – with high resolution and low relief – is unique today and is called "2.5D", not 3D. To date, the equipment has been successfully tested on engravings, cartographic materials, seals, fabrics and other objects of museum storage.

However, the second leading trend of the 2010s after the gigapixel scanning technology was the use of artificial intelligence (AI, Artificial Intelligence, AI) to solve various research tasks, which was confirmed by a series of seminars held in 2019 by the professional network community "Museums+AI" (URL: https://themuseumsai.network /) and the report published in February 2020 about the possibilities of using AI in museums ^[167].

Perhaps the most famous project in this direction is the restoration of Rembrandt's painting "Night Watch", stored in the Amsterdam Rijksmuseum Art Museum. In the course of the work, which was covered on a special website (URL: https://www.rijksmuseum.nl/en/stories/operation-night-watch ), using AI techniques, two gigapixel images were created. The first was published in May 2020 . The copy size was 44.8 gigapixels. Just a year later, this record was surpassed by the researchers themselves, and the largest and most detailed electronic copy ever made from a work of art was presented on the museum's website. Its size is 717 gigapixels, the distance between neighboring pixels is 5 micrometers (0.005 millimeters), and the total file size is 5.6 TB. (URL: https://www.rijksmuseum.nl/en/stories/operation-night-watch/story/ultra-high-resolution-photo ). To create this image, the researchers used a 100-megapixel Hasselblad H6D 400 MS camera, which was used to take 8439 individual photographs measuring 5.5 x 4.1 cm. The subsequent integration of the obtained fragments into a single image was carried out using artificial intelligence algorithms.

The restoration work carried out in this project using the most modern information technologies and equipment allowed the researchers to implement several scientific developments, including:

- obtaining images using a macro X-ray fluorescence scanner, showing the presence of different colors and pigments in the picture (in particular, dark blue cobalt), which are not visually detected (URL: https://www.rijksmuseum.nl/en/the-first-scan );

– virtual reconstruction of a full-size Rembrandt painting (URL: https://www.rijksmuseum.nl/en/stories/operation-night-watch/story/artificial-intelligence ) ^[168];

- presentation of research results on the Internet (https://www.rijksmuseum.nl/en/whats-on/exhibitions/operation-night-watch/story/results-of-the-research ) based on the tools and standards of the International Image Interaction Platform (IIIF – URL: https://iiif.io /), which allow virtually without loss of quality, imperceptibly and smoothly combine and present image fragments (using the API) stored physically remotely from each other; segment single images and store they are scattered; describe/comment/annotate each fragment, forming databases, and use other tools for working with electronic images ^[169].

Ideologically similar projects of using high-quality electronic images of museum objects for the purposes of digital restoration and/or study using DS, machine learning, artificial intelligence, computer vision, etc. are being implemented in Austria (digital restoration of the coloristic solution of Klimpt panels ^[170]); USA (digital restoration of the Dutch still life of the XVII century ^[171], attribution paintings "Savior of the World" ^{[172; 173]}, study of the flaps of the polyptych of the Van Eyck brothers (Ghent altar – URL: http://closertovaneyck.kikirpa.be ) ^[174]; development of technology for establishing the authorship of canvases with a probability of up to 96% ^{[175; 176]}, etc.

At the same time, the use of AI for the study of museum objects, and in particular, objects of painting, is not so unambiguous. For example, in the fall of 2021, a discussion article by a specialist in medieval fine art, Sonya Dreamer, was published in the professional online publication The Subversion about whether AI really helped in the study of the Ghent altar, the restoration of Rembrandt's "Night Watch", Picasso paintings, etc., as various publications and experts write about it, relying on statements by employees of Oxia Palus, who took part in the work on these projects, or is it just a distraction, covering up less benign areas of AI application with ambitious projects ^[177]. It seems that the concerns expressed by the author are not unfounded, and the question of the value of AI as a method of research in the field of art history and art criticism is formulated quite rigidly and requires reflection.

The authors of this article adhere to cautious assessments and believe that AI still makes it possible to intensify the use of many traditional methods of technical and technological analysis, and the development of original research methods for studying works of art using AI is a matter of the future.

The third trend of the 2010s in the field of working with images – the development of visual search platforms in historical and cultural heritage collections - is being actively implemented worldwide (for example: Art–UK; Japanese Woodblock Print Search; PHAROS; as well as Google Arts & Culture initiatives, etc.). So, in the spring of 2020, the UK launched The Deep Discoveries Project (URL: https://tanc-ahrc.github.io/DeepDiscoveries/index.html ), the purpose of which is to develop a "visual search" based on the use of one of the methods of artificial intelligence: the comparison of similar images taking into account their visual characteristics (color, pattern, shape). The project aims to create a computer vision search platform that can identify and compare images on the scale of all digitized collections stored in England. A similar project is being implemented by the staff of the Norwegian National Museum ^[178].

And a group of researchers from Microsoft and the Massachusetts Institute of Technology (MIT) in the MosAIc project (project presentation — URL: https://microsoft .github.io/art /) has complicated the task and is creating a system for finding relationships between art objects from different cultures and eras, which allows you to track the process of cultural exchange between different territories in time (URL: https://microsoft .github.io/art/app/%3Fid%3DMTk0MTEz).

In Russia, the use of specialized technologies for obtaining high-resolution electronic images is still concentrated in the field of restoration and technical and technological study of museum objects. However, in the last decade, a few joint works of museum staff and specialists of scientific organizations specializing in modern methods of multispectral macro photography and information technology have begun to appear. Several similar studies have been carried out by employees of the GIM, the Space Research Institute of the Russian Academy of Sciences and the National Research Center "Kurchatov Institute" ^{[179; 180]}.

Descriptions of projects that have already been implemented or are currently being carried out can be continued, but it seems that the examples given are enough to confirm the correctness of the conclusion made by the task force of the European project EuropeanaTech AI for GLAMs and published in autumn 2021 in the analytical report on the main directions of AI application in museum activities ^[181], in which as the most technologies for creating high-quality electronic images and working with digitized collections, including using Data Science methods, are currently in demand for the museum community.

Summing up the review of the history of the development of systems for creating electronic images of museum objects and their (electronic images) application in intramuseum activities (in the field of restoration, conservation, cataloging, scientific research, publication, etc.), the authors summarized in one table (Table No. 1, Part 1-3) criteria that served as the basis for the allocation of chronological periods, and the results achieved during each of them. The content of the Table cells clearly demonstrates the progressive development of technology, technologies and methods of using electronic images, in which at each subsequent stage of evolution the achievements that were made at the previous ones were preserved, new methodological, technical and technological solutions appeared, allowing to expand the range of museum objects for which electronic images are created, as well as to improve the methods of their study.

Table No. 1. The evolution of technology and technology for creating electronic images of museum objects, compression formats, and the purposes of using EI.

Tsch. 1: Technical and technological aspects

Table No. 1 . The evolution of technology and technology for creating electronic images of museum objects, compression formats, and the purposes of using EI

Tsch. 2: Digitization objects, organizations, projects Table No. 1 . The evolution of technology and technology for creating electronic images of museum objects, compression formats, and the purposes of using EI.

Tsch. 3: The purpose of creating and using electronic images; user audience

Concluding the presentation of the results of the conducted historiographical research, the authors consider it necessary to emphasize that:

- the historiographical problem posed is far from being exhausted and deserves a more in-depth study and description in the historical context, both on the basis of the involvement and analysis of a wide range of already identified sources, and with the involvement of the results of monitoring and research of current trends and practices implemented in various projects;

- the proposed periodization is of a debatable nature and can be adjusted under the influence of newly revealed facts and scientific publications.

In any case, the study of the digital transformation of the methods of technical and technological analysis of museum objects, which occurred in the second half of the XX - early XXI centuries, as well as the study of the history of the design and development of equipment designed to create high–quality electronic images, specialized software applications and the introduction of Data Science methods into intramuseum activities (restoration processes, attribution, authenticity examination ensuring the safety and security of museum collections, art history research, etc.) seems to be an extremely promising direction, and knowledge of historiography, content, technical and technological solutions, results and conclusions obtained in the most significant projects can help researchers optimize further scientific search.

Notes

1. "In order to preserve cultural heritage ... we need to be able to fully recreate the object of historical and cultural heritage in digital form, using various digitization technologies, as reliably and fully as possible, and even predict what will happen to it over time. And even better - to observe the object, including in the storage, to watch for changes in its state. For some exhibits, we are still constantly conducting such partial monitoring with a frequency of once every six months or less." Cit. by: Olga Melnik. Interview: "Vladimir Definenov: "The biggest difficulty is that we are trying to digitize analog processes" // IT World. 31.03.2021.

It is obvious that V. Definenov confuses the 3D model of the object and the database for managing the originals of museum objects, where the model can be integrated as an illustration. The reflection in the model of changes occurring in the object itself in the process of its existence (including in the museum) (condition monitoring) is carried out either by re-creating and updating the model (which is a rather laborious and capital-consuming process, by no means safe for the preservation of the object itself), or on the basis of information documented by the keepers. In any case, the idea is not new, it has already been implemented in the EROS system described above and worked out in detail in the Program of informatization of the accounting and stock work of the GIM.