Using Digital 3D Models for Study and Restoration of Cultural Heritage Artifacts (Digital Imaging) Part 1

Introduction

The introduction of new technologies in the context of cultural heritage (CH) and Archeology has often been a difficult issue. This is probably related to the lack in confidence in replacing consolidated approaches with experimental methods heavily based on innovative hardware or software systems. This already happened for a number of revolutionary technologies: for example, the advent of photography, color images and digital cameras took some time before changing the reference methods for archival and studies in the context of archeological excavation or of restoration actions. The same considerations hold for the use of digital 3D models in CH applications. One basic issue is the need to switch from a two-dimensional visualization and reasoning approach (essentially based on photos and drawings) to the possibility to explore and visualize the object in its full three-dimensional (3D) nature. Nevertheless, in the last few years both 3D modeling and 3D scanning have become a valued way to present and analyze CH artifacts. Several interesting practical applications have been made available to the public in museums, or virtually on the Web. Here we just cite a single sample paper [56], which deals with the problem of interactive inspection and rendering of complex 3D models; many other experiences are presented in this topic.

The availability of tools for the reconstruction and the virtual interaction with accurate digital 3D models, supporting accurate and flexible visualization features, is a crucial issue for many applications, such as architectural design, graphics simulation, and scientific visualization. The users of these systems need accurate, realistic and/or interpretative visual representations, real-time navigation, and flexible interaction tools. Ease of use is also an important issue in the design of visualization tools for CH applications, because most CH users (such as museum curators, art historians, restorers, and museum visitors) have limited skills in managing 3D graphics technology. The graphics user interface (GUI) design and the overall usability of the rendering tools play a critical role in determining whether the tool will be just a nice toy or a useful technical instrument. Another problem is that the complexity (the huge number of graphics primitives required to represent a shape) of accurate, realistic-looking models exceeds the interactive capabilities of most graphics workstations. We need real-time visualization of those huge meshes without sacrificing the high-quality achievable with 3D scanning technologies. We have several examples of use of 3D models in virtual presentation of CH artworks to the wide public (just to cite a few, see [2-6]), most of them very successful in their field. On the other hand, visualization to the public is only one of the possible goals. Another important objective is to design applications or tools to help CH experts in their everyday work, being it the study or research on CH, or the conservation/restoration. Until now, the use of 3D graphics and 3D scanning in CH applications has been limited by the costs of the acquisition devices and the slowness and complexity of the processing phase required to transform raw sampled data into accurate and usable digital 3D models. Technological issues play a role, but this is not the only (or major) reason for the scarce use of 3D graphics in CH applications. Another key factor is the lack of some killer applications in the realm of the daily needs of CH experts. Visualization is the most diffuse use of digital 3D models, but unfortunately in many cases virtual 3D presentation is still considered by CH practitioners as a medium to reach the public, rather than a basic instrument for CH research and conservation. This topic aims to show that there are wide possibilities for the use or development of software tools oriented to the CH domain. This activity is often perceived by the academic community as an applied research or a research transfer activity (designing tools according to specific needs of the CH users community). This is partially true, since in some cases academic research and development is playing the role that should usually be covered by commercial companies; but the small budget that characterizes most of the ICT application in the CH domain can drastically change in the near future, raising the level of interest of the industry. On the other hand, in many cases the specific CH needs motivated interesting basic research on CH technology and had an impact on the advance of our research domain. Some of the possible applications of digital 3D models in the CH domain together with a description of some tools specifically designed for CH applications are presented in the following, trying to cover several potential interests of CH experts.


Visual Communication of Art

As briefly mentioned in the introduction, visual communication is by large the most common utilization of digital 3D models in the CH domain. Just to make an example, the 3D media is becoming more and more a common resource in the production of broadcasted programs on history and art. The different visual presentation modalities can be divided in those using still images and those adopting video streams. The latter can be further divided into passive modalities (i.e., videos or computer-generated animations) and active modalities (i.e., interactive multimedia installations or virtual navigation systems.) Let us briefly present the potential of those technologies, and some major experiences and issues.

Some snapshots from “The Parthenon” .

FIGURE 2.1

Some snapshots from “The Parthenon” .

Computer-Generated Animations

Computer-generated animations are by far the more common channel to present digital 3D models or hypothetical reconstructions of CH artworks to the large public (Figures 2.1 and 2.2). Nowadays, any broadcasted program on art or history contains at least some animation clips; high-budget movies recreate the past using the available technology at its best. The Hollywood blockbuster “The Gladiator” is just one example of the many historical movies which allow to visually present highly veridical representations of our past. There is a long standing discussion on how far these representations are also a good medium to convey historically or culturally veridical representations of the past (i.e., up to what extent they comply with art history knowledge). Technology is nowadays extremely high quality and the images/videos produced are so compelling (photorealistic), that the resulting video sequences can become the historical reality in the mind of the public, maybe much more than that reached/obtained by years of CH research and dissemination. But, as usual, the problem is not in the medium, but in its use. One rising issue is that the level of quality reached by the current movie production is paired by an extreme complexity of the production process and of the related costs. Computer animation systems are now very complex, flexible, and expensive. They also raised the bar of the public perception of what is a good-quality virtual representation or experience. Anyone attempting to produce a video or a visual presentation (for example in the CH domain) should expect the potential users will make their personal evaluation by comparing it to what they are used to seeing (i.e., Hollywood movies). It is often very hard to run against such tough competition.

Nevertheless, there have been very nice examples of videos produced in the academic/research domain, based on a low budget and giving very good examples of what can be accomplished when we pair CH dissemination needs and new visual technologies. One of the best videos ever produced in this domain is “The Parthenon” movie, produced by Paul Debevec(http://www.debevec.org/Parthenon/film.html) in 2004. The video is a very good example of how the story of a monument can be presented; it visually reunites the Parthenon and its sculptural decorations, separated since the early 1800s. The film used a number of cutting-edge technologies (3D scanning, photometric stereo, inverse global illumination, photogrammetric modeling, image-based rendering, BRDF measurement, and Monte-Carlo global illumination) in order to create the twenty-some shots used in the film.Another more recent example of intense utilization of computer animation to present CH artworks is the video content at the base of “The Virtual Museum of Iraq” (http://www.virtualmuseumiraq.cnr.it/). Most of the animations presented there have been produced by an Italian CNR group, led by Francesco Gabellone, using mostly image-based reconstruction approaches and commercial video production technology. Those animations can also be considered as a nice example of what can be produced by a small multidisciplinary group (computer scientists, architects, and art historians) working under a small budget.

Snapshot from the movie clips presented on “The Virtual Museum of Iraq” .

FIGURE 2.2

Snapshot from the movie clips presented on “The Virtual Museum of Iraq” .


Computer animation can be an ideal instrument as well for visualizing the CH scholars’ research results, if the production complexity and cost will reduce. We have indicators that the diffusion of the knowledge on computer animation technology is considerably improving in young generations. Another important factor for a wider diffusion is the impressive improvement of low-cost open source software platforms (e.g., Blender [7]).

Interactive Visualization

An advantage of interactive visualization is to insert the user in the loop. Conversely to passive media such as computer animation, it is now the user that drives the navigation and the inspection of the digital artifact. An interactive system allows the user to follow his specific interest while choosing the exploration path, focusing on the details that hit personal interest and giving the possibility to choose the duration of the visualization session on the base of the specific insight experience and needs. Virtual Inspector (see Figure 2.3) [56] is an example of a CH-oriented visualization system which aims to fulfill some of the issues sketched before. The system allows naive users to inspect a very dense 3D model at interactive frame rates on off-the-shelf PCs, presenting the 3D model and all the multimedia data that has been linked to selected points of its surface. A main goal in the design of the system is to provide the user with a very easy and natural interaction approach, based on a straightforward point and click metaphor. Visualization efficiency is obtained, without sacrificing quality, by adopting a state-of-the-art continuous level-of-detail (LOD) representation [68]. Finally, the adoption of the XML encoding for the specification of the GUI structure and behavior makes Virtual Inspector a very flexible and configurable system.

A snapshot of the Virtual Inspector interface.

FIGURE 2.3

A snapshot of the Virtual Inspector interface.

The Arrigo Kiosk, with Virtual Inspector integrated in a website-like kiosk.

FIGURE 2.4

The Arrigo Kiosk, with Virtual Inspector integrated in a website-like kiosk.

Virtual Inspector has been used for the implementation of several multimedia presentation, including: the Arrigo VII kiosk [9] (Museo Opera Primaziale Pisana, on show since the end of 2004, see Figures 2.3 and 2.4), a system for documenting the results of the restoration of Michelangelo’s David [10], an interactive kiosk to present the excavations results in the area of Luni (La Spezia, Italy), etc. But Virtual Inspector is just an example of those systems devoted to the inspection and manipulation of a single artifact. The virtual navigation of large scenes, representing a complex of artifacts such as for example a complete archaeological site or a historical town, is also of great interest. The need to represent and navigate a large scene opens a number of issues [11] that have to be considered, namely:

•    Need to sample a large extent, which has to be processed to build up the global model (processing huge 3D sampled domains is not as easy as processing a single statue);

•    Representing the 3D model in a format which is adequate for interactive visualization (the adoption of multiresolution methodologies for encoding and rendering the 3D models is here mandatory);

A snapshot of the VR installation in Ename, Belgium [12].

FIGURE 2.5

A snapshot of the VR installation in Ename, Belgium [12].

• Devising intuitive GUI approaches for driving the navigation in a complex scenario; usability is a must, to prevent that the user would be concentrated on how to drive the navigation rather than enjoying and getting insight from the virtual visit.

Moreover, the use and combination of cutting-edge technologies can boost the user experience also in the visit of large areas: the example of Ename [12] shows the combination of several approaches for the comprehensive presentation and exploration of an interesting historical settlement (Figure 2.5).

Geographic Web Browsers Deploying 3D Models

Web-based systems such as Google Earth [13] or Microsoft Live Local [14] could boost considerably the usage of 3D reconstructions and the availability of 3D models of CH artifacts. These systems allow the addition of small-resolution 3D models on top of terrain models, and link those representations with high-resolution 3D models of the architectures and artworks which characterize a given territory or urban context. We could easily forecast that in the near future local authorities will be active sponsors for massive data gathering to populate those systems, aiming at increasing the tourism industry and providing a wide dissemination of local CH masterpieces. This could give an impressing impulse to the use of both image-based approaches, for the reconstruction of low-polygon models, and 3D scanning, for the production of higher accuracy models.

Open source platforms are also available for large-scale representations of CH, supporting virtual navigation of CH resources, either on site or on the web. Some applications have been experimented by the Virtual Heritage lab of CNR-ITABC, focusing on the reconstruction of archeological sites. An example is the “Virtual Museum of the Ancient via Flaminia” [15] (http://www.vhlab.itabc.cnr.it/flaminia/). The system has been designed as a Virtual Reality system to be installed inside an important museum in Rome. One of the design goals was to keep the social aspects which are typical of the visit in a museum, therefore it was designed to support a multi-user navigation and interaction. Through the VR application, visitors can explore the landscape that characterized the Roman road, Via Flaminia, and have an intense experience visiting four different sites (an old bridge over the Tevere River, a Roman necropolis, the Villa of Livia, wife of emperor Augusto, and finally a Roman arch), all of them reconstructed in details using different 3D technologies. Authors performed a topographical, archaeological, and architectonic survey, whose goal was data acquisition at different levels of detail (DGPS, laser station, 3D laser scanner, photogrammetry, cartography, remote sensing, aerial and satellite photos processing). Data have been integrated in the system, to build up models of the landscape and of the CH artifacts, which could be navigated in interactive time with the selected VR platform. A similar navigation experience is also available (in single-user mode) on the web via the support of the VR Web GIS platform [16].

Next post:

Previous post: