Journal of Cultural Heritage 43 (2020) 129–143
Available online at
ScienceDirect
www.sciencedirect.com
Original article
3D reconstruction and validation of historical background for
immersive VR applications and games: The case study of the Forum of
Augustus in Rome
Daniele Ferdania,∗
, Bruno Faninia
, Maria Claudia Picciolia
, Fabiana Carbonia
, Paolo
Vigliarolob
a
CNR ISPC, via Salaria KM 29300, 00015 Monterotondo St. Rome, Italy
b
Sovrintendenza Capitolina ai Beni Culturali, Mercati di Traiano – Museo dei Fori imperiali, via IV novembre 94, 00187 Rome, Italy
a r t i c l e i n f o
Article history:
Received 2 October 2019
Accepted 10 December 2019
Available online 10 January 2020
Keywords:
3D source-based reconstruction
Virtual reality
Virtual archaeology
Serious game
VR game
Head-mounted display
Historical validation
Virtual museum
Edutainment
a b s t r a c t
In the last decades, thanks to the success of the video games industry, the sector of technologies applied to
cultural heritage has begun to envisage, in this domain, new possibilities for the dissemination of heritage
and the study of the past through edutainment models. More recently, experimentation in the ﬁeld of
virtual archaeology has led to the development of virtual museums and interactive applications. Among
these, the “serious game” segment – the application of interactive technologies to the cultural heritage
domain – is rapidly growing, also including immersive VR technologies. Applied VR games and applications
are characterized by a thorough historical background and a validated 3D reconstruction. Indeed,
producing such products requires a tailored workﬂow and large effort in terms of time and professionals
involved to guarantee such faithfulness. Drawing on our previous work in the ﬁeld of virtual archaeology
and referring to recent experiences related to the deployment of applied VR games on PlayStation VR
©, we describe and assess a workﬂow for the production of historically accurate 3D assets, targeting
interactive, immersive VR products. The workﬂow is supported by the case study of the Forum of Augustus
and different output applications, highlighting peculiarities and issues emerging from a multi and
interdisciplinary approach.
© 2019 Published by Elsevier Masson SAS.
1. Introduction
The use of virtual reality (VR) within the cultural heritage
domain dates back to the end of the 80s and Paul Reilly was the
ﬁrst researcher to coinage the term “Virtual Archaeology” (VA) in
1990 to describe the use of computer-based simulations of archaeological
excavations:
What does the term virtual archaeology mean here? The key concept
is virtual, an illusion to a model, a replica, the notion that
something can act as a surrogate or replacement for an original. In
other words, it refers to a description of an archaeological formation
or to simulated archaeological formation” [55].
∗ Corresponding author.
E-mail addresses: daniele.ferdani@cnr.it (D. Ferdani), bruno.fanini@ispc.cnr.it (B.
Fanini).
URLs: http://www.ispc.cnr.it/ (D. Ferdani), http://www.itabc.cnr.it (D. Ferdani).
From the ﬁrst experiments, an onward interest has arisen
towards a thoroughly combination between humanities and hard
science in order to achieve reliable reconstructions of the past.
Since the 90s, method, theory and application of three-dimensional
simulation and visualization in cultural heritage have been annually
debated in several conferences (e.g. CAA, Eurographics, Digital
Heritage) and internationally recognised principles have been
established by the scientiﬁc community such as “London Charter”1
or “Sevilla Principles”.2 According to the most recent studies we
could deﬁne the VA as a process that allows to reconstruct and
simulate the past (artifacts, architecture or landscapes) through a
theoretical and multidisciplinary approach and the use of digital
technologies [2,51].
Continuous experimentation and cross fertilization between
the disciplines, led the community to new horizons of research.
On the one hand they opened new perspectives and possibilities
1
www.londoncharter.org.
2
www.sevilleprinciples.com.
https://doi.org/10.1016/j.culher.2019.12.004
1296-2074/© 2019 Published by Elsevier Masson SAS.
130 D. Ferdani et al. / Journal of Cultural Heritage 43 (2020) 129–143
of analysing, interpreting and simulating the past, on the other
hand they contributed to develop speciﬁc technologies or adapt
paradigms and tools, originally created for other ﬁelds of applications
(computer graphics, VR, image-based modeling, etc.), to
this domain. More recently, researchers paid close attention to
those technologies, media and communication paradigms aimed
at heritage fruition through 3D real-time applications such as virtual
reality, augmented reality, serious game, etc. They could allow
users to experience history and archaeology both visually and narratively,
to observe and interact with artifacts, architecture and
characters belonging to a speciﬁc period in history. Historical content,
in fact, conveyed by means of this media with a scientiﬁc
approach, could help institution in dissemination and understanding
of the past [3,52,4,11].
The start was slow and hesitant due to the absence of adequate
technological solutions and theoretical guidelines. Furthermore the
information science still had to formulate appropriate communicative
languages and paradigms for this emerging domain. Given
that, the ﬁrst experiments were too complex, lacking of interest
for users or deﬁcient of historical accuracy and validity. Nowadays,
thanks to the evolution of effective theories, methods and tools, virtual
archaeology and 3D application for heritage are increasingly
growing inside museums and sites. This success depends also on
the use of dramatization and storytelling which take advantage of
emotional processes to engage users and disseminate knowledge
[54,47].
One of the most innovative ways for understanding culture and
history is the “serious game” or rather the adaptation of an interactive
technology, well-known in the entertainment industries, to
the cultural heritage domain [5].
1.1. Games and culture: state of the art
The success of entertainment game industries has given birth
to new types of outputs in the ﬁeld of cultural heritage, including
serious games and some virtual museum applications (based
on mixed reality, virtual reality, virtual worlds, etc.) that share
the same infrastructure and core games technologies and use virtual
reconstruction and engagement mechanism for edutainment
and educational purposes [1]. Ancient Pompeii [46], the Virtual
Museum of the Ancient Via Flaminia [32], the Virtual reality reconstruction
of Otranto [16] or Roma Nova [64] – virtual experiences
which allow the exploration of ancient contexts – are just some
of the ﬁrst examples of interactive applications for the cultural
heritage with game-like navigation mechanism.
More recently, serious games, thanks to its intrinsic balance
between learning, narration and fun experience, are becoming very
popular. A virtuous connection between games and culture is constantly
growing and its impact on the society is testiﬁed by some
recent virtuous case studies which uses storytelling and modern
video games logic [12,62,47].
In 2011 the Foundation of the Hellenic World produced Battle
of Thermopylae, a game that mixes storytelling techniques and
game to disseminate historical knowledge about the battle and the
associated legends [15].
In 2012 the Department of Art History and Visual Studies of the
Duke University performed the Fort Ross Virtual Warehouse project,
a 3D real-time multiplatform serious game aimed at discovering an
outpost located on a bluff along the Northern California Cost and its
landscape with the purpose of disseminating and teaching cultural
information related to the site in an innovative way [42].
In 2015 Defend the Wall was created by Evoca srl. It is serious
game that deals with the technology and tactics of siege of the
4th Century BC, the war machines and the strategies of attack and
defense, the architectural characteristics of the towers and the city
walls of Paestum. The issue is developed through a short movie
followed by a game using gesture-based interaction, suitable for a
diverse audience, particularly young people [31].
In 2017 the MANN, Archaeological National Museum of Naples,
produced Father and Son,3 a narrative 2D game rated 4.7/5 on
Google Play and downloaded by more than 33,000 players. It is
a story of a son who travels across time and space, from the ancient
Rome and Egypt to the modern Naples, to discover the life of his
father, an archaeologist he never knew.
In 2018 was released Mi Rasna – Io sono l’Etrusco, a strategy game
for mobile based on the Etruscan civilization where the user discover
the ancient society and earn extra-bonus when geo-located
on Etruscan archaeological sites or museum.4
During the last years VR games are emerging in the cultural
panorama. In 2018 the start-up “Beyond the gate” produced the
on-site VR game called Beyond the Castle. The game is located in the
ancient Torre Falconiera of the Sforzesco Castle (Italy), where visitors,
wearing HDM and taking up a special crossbow, have to defend
the castle under siege.5 The current Horizon 2020 ﬁnanced project
i-MARECULTURE6 investigates immersive technologies – virtual visits,
serious games with immersive technologies and underwater
augmented reality – to increase exploration time in an underwater
archaeological site as in the case of the VR visualization of the
Mazotos shipwreck [43].
An interesting application is Arkaevision Arkeo. It is an immersive
VR application aimed at discovering the temple of Hera in
Paestum Posidonia. During the experience, the user ﬁnds himself
interacting with a digital actor, the priestess Ariadne, who accompany
him on an interactive journey through the rituals, myths and
architectural and artistic artifacts of the Magnogreek city in the 5th
Century BC [9]. This application was developed within Arkaevision,
an integrated digital platform which uses immersive virtual reality
and mixed reality as main technologies and experiential modes for
disseminating cultural heritage.7 It is the result of a research and
development project funded by MISE within “Horizon 2020 Call for
Proposals” and was carried out by Digitalcomoedia in collaboration
with CNR ITABC and Beyond.
All the mention games have been designed and developed by
specialist in collaboration with institution (museum or archaeological
sites) and under the scientiﬁc assistance of expert in different
cultural domain. Peculiar examples are the Discovery Tour of Origins
and Odyssey, the last two chapters of the Assassin’s Creed saga,
produced by the famous Ubisoft Entertainment game industry.8
Discovery tour is a purely educational game mode that allow users
to explore and interact with reconstructed historical sites through
guided tours divided into thematic areas and according to recent
studies, performed by the V-Must consortium on digital heritage
sectors, this spin-off application can be properly considered as a
virtual museum.9 The great novelty of this two last chapters is the
innovative approach and attention the creators paid to historical
reconstructions. Indeed Ubisoft entrusted a team of historians and
archaeologists with the role of consultant for supporting the team
and validating virtual reconstructions.
In this kind of games, which take place in a given and well known
historical period, three-dimensional assets and virtual reconstructions
take on a fundamental role. They are the setting for the
game and the success depends not only by the screenplay but also
by the accuracy and the faithfulness of the represented histori-
3
www.fatherandsongame.com.
4
www.egameapps.com/mi-rasna.
5
www.beyondthegate.io/en/homepage.
6
www.imareculture.eu.
7
www.arkaevision.com.
8
www.ubisoft.com.
9
www.v-must.net/virtual-museums/what-virtual-museum.
D. Ferdani et al. / Journal of Cultural Heritage 43 (2020) 129–143 131
cal environment. Both historic and imaginary world are perfectly
integrated. Maxim Durman, historian of the Ubisoft team for Origins,
during an interview said that even if playing the game cannot
substitutes historical studies, however it plays an important role
in edutainment, supporting didactic activities and disseminating
information through amusing experiences. For this reason any presented
asset in the historical reconstruction must be accurate as
much as possible10 .
Projects oriented at integrating heritage and video games are
consistently increasing also in the ﬁeld of applied science. Creative
Europe and Horizon 2020 are focusing on this topic and the recent
announcements are even more oriented at edutainment, fostering
creative industries, researchers and museums at developing this
media in museum in order to improve visitor engagement, interaction
and learning mechanism. Interesting results on advantages of
applied immersive VR games compared to traditional approaches,
were already obtained for instance in [9,59,49] and [40].
2. Research aim
Historical accuracy and validation are the keywords that portray
the virtual backgrounds made for applied VR games and applications
mentioned in the previous sections, however in order to
guarantee such faithfulness their production requires a complex
and well-founded workﬂow. Indeed, creating such interactive VR
applications targeting cultural heritage requires extra effort in
terms of time and professionals involved which rigorous collaboration
is crucial to achieve robust results. The aim of this work is
to provide a complete reproducible reconstruction workﬂow – purposely
designedfor real time VR products,highlightingpeculiarities
of this multi and interdisciplinary work and challenges arisen, with
practical application to the case study of the Forum of Augustus.
3. Material and methods
3.1. The case study
The case study we are going to deal with in order to better
illustrate the workﬂow is the 3D reconstruction of the Forum of
Augustus at the beginning of the 1st Century AD for VR cultural
applications and games. In Section 4 this aspect will be dealt with
in detail.
The Forum of Augustus is a magniﬁcent architectural complex
built by the Emperor Augustus between the end of the ﬁrst century
B.C. and the beginning of the ﬁrst century A.D. and it is the
second of the Imperial Forums of Rome, after the ﬁrst forum built
by Caesar (ﬁg. 3). It was a monumental complex which hosted the
majestic temple of Mars Ultor (the Avenger): Augustus made the
vow to build that temple before the battle of Philippi (42 B.C.),
taking revenge on the murderers of Julius Caesar, Brutus and Cassius.
Two porticos supported by corinthian colonnade enclosed the
square, and behind this columnade there were four semi-circular
exedrae (two for each side of the forum). The two major exedrae
could have served as courts while the two minor were probably
used as archives and for educational activities (Gymnasium). The
back wall of the porticoes and of the exedrae were articulated with
columns framing a series of rectangular niches adorned with a rich
variety of different statues. At the end of the northern portico, there
was the Hall of the Colossus, a decorated chamber which hosted the
colossal statue, approximately 11 m tall, representing the Genius
Augusti (the protective deity of Augustus).
10
Interview on http://www.rollingstone.it/cultura/gaming/page/2/ (accessed
13.02.19)
Nowadays, the remains of the Forum are still visible and well
preserved in the archaeological area of the Imperial Fora (see Fig. 1),
and all the ﬁndings discovered during the excavation performed
during the last century (statues, architectural elements and decoration)
are shown in the nearby museum of the Trajan’s Markets –
Imperial Fora Museum. Even if the Forum of Augustus is one of the
most known and studied buildings from Roman times and several
reconstructive hypotheses were drawn over the years, archaeological
discoveries have revealed new elements and new reasoning,
and hypothetical reconstructions have been proposed and adopted
in this work [60,61,41].
Our laboratory has already designed a full 3D reconstructions
of the Forum for “Admotum” [27], a gesture-based application created
for the exhibition “Keys to Rome”11 which was held between
2014 and 2015 at the Museo dei Fori Imperiali and organised by
V-Must, the virtual museum Transnational Network.12 This reconstruction,
however, represented the Forum in the Trajan’s period
(98-117 A.D.) when its shape had been signiﬁcatively changed as
a result of construction work carried out for the ediﬁcation of the
Forums of the emperors Nerva and Trajan. In addition, some internal
areas of the forum had not been modelled in detail as they were
not accessible in the previous application. Given that, it was decided
to restart from this baseline by remodelling the Forum from scratch
according to an approach more oriented to the world of games but
reusing, when possible, the 3D assets and historical study.
3.2. Preliminary remark
3D models created in the reconstruction of the game scenarios
are commonly deﬁned as “ﬁgurative”, which means they represent
reality in a plausible way. Although their complex topology is
optimized to respond to operational needs, they do satisfy some
important characteristics such as the recognizability of the basic
geometry that composes the object and the characterization of the
surfaces with chromatic information derived from photographs or
procedural images (textures) [7]. The virtual reconstruction of the
Forum of Augustus was created on the basis of direct surveys –
characterized by great precision and reliability – study and interpretation
activities based on scientiﬁc approaches and principles
of virtual archaeology [14,44]. This operation has guaranteed the
historical reliability of the model but required an articulated and
multidisciplinary workﬂow that we have divided into three main
segments and can be reused for future project (NB: The workﬂow
described below refers only to the creation of cultural assets and is
not exhaustive of all the development phases of a video game for
which we ask the readers to rely on speciﬁc literature):
• Pre-production (Section 3.3): preliminary studies including 3D
survey and digitisation activities and historical sources collection.
• Production (Section 3.4): it has mainly concerned the aspect
of archaeological interpretation, reconstruction through 3D
modelling of game assets, validation of historical background,
performance analyses and optimization of the assets.
• Level creation and authoring (Section 3.5): it was mainly connected
to the graphic layout and environmental simulation through the
composition of the scene within the graphics engine and the addition
of details that confer realism, such as lights, particle effects,
sounds, etc.
11
www.keys2rome.eu/eng/.
12
www.v-must.net/.
132 D. Ferdani et al. / Journal of Cultural Heritage 43 (2020) 129–143
Fig. 1. Right: the archaeological site of the Forum of Augustus nowadays. Left: Hypothetical reconstructive layout of the Forum in the Augustan era overlaying the archaeological
remains (grey area). (A) Temple of Mars Ultor; (B) Hall of Colossus; (C) Guardpost; (D) Major exedrae used as tribunal; (E) Minor exedrae used as archives; (F) Porticos;
(G) Central open area.
3.3. Pre-production workﬂow
This phase, which is preparatory to the creation of the game
assets, is aimed at collecting all the documentation useful for the
historical reconstruction of the game context. The documentation
can be dived in two categories:
• geometric acquisition of cultural site and museum objects using
image-based or range- based sensors with the aim of obtaining
digital replicas.
• collection of all the existing historical and archaeological sources
and testimonies regarding the sites and all the connected artifacts
and testimonies.
3.3.1. 3D survey and digitisation
Technologies and tools for digitization are numerous: from
three-dimensional surveying (laser scanner and digital photogrammetry)
to topographic surveying, each with different characteristics
such as accuracy, precision, portability, cost and automation [56].
The models, built on real data, ensure a high degree of realism in
terms of formal coherence and aesthetic rendering [38].
During the 3D survey of the Forum of Augustus we used ImageBased-Modelling
(IBM) method which combines computer vision
and photogrammetric algorithms to obtain three-dimensional
models from a dataset of two-dimensional images taken around
the artifacts. In this particular case we used a full-frame camera
equipped with a 35 mm lens to capture photos and a laser total station
to measure the topographic network of control points (which
are fundamental to achieve models featuring high geometrical precision
and georeferencing accuracy). Using this method we scanned
both the architectural remains of the Forum, in Via dei Fori Imperiali,
and the remains of the marble decoration of the porticoes and
the Temple of Mars Ultor preserved in the Museum of the Imperial
Forums.
The data gathered on the ﬁeld were processed using Agisoft
Photoscan13 software. The software was able to estimate camera
positions and then calculate depth information for each camera
to be combined into dense point clouds. Subsequently the point
clouds were triangulated to obtain meshes. Last step for creat-
13
www.agisoft.com/.
ing the virtual replica of the scanned objects was the automatic
parametrization of the meshes and the generation of textures. The
ﬁnal result of the digitisation process can be seen in Fig. 2.
3.3.2. Sources collection
It concerns the collection of all available historical documentation
and archaeological evidences that can support the virtual
reconstruction. The reliability of the reconstruction is determined
by the meticulousness of the work of collecting and managing these
data [34,33]. The sources used for the interpretation and 3D reconstruction
can be uncountable. It depends on the context we are
dealing with. In our case we have used the following:
• Extant structures: archaeological evidence still visible in situ.
• Archaeological evidence removed from the site whose original
position is known.
• “special ﬁnds”, archaeological evidence still existing but non
in situ (i.e. found outside their original context) such as architectural
fragments of lintels, columns or roof tiles collapsed to
the ground (their existence, appearance and size is certain, while
their original position is uncertain).
• Testimonies from the past, such as descriptions of buildings and
their function contained in historical documentation; literary
sources; ancient iconography; old photos, drawings, paintings,
watercolours illustrating previous state of conservation; technical
drawings from the 20th century.
• Scientiﬁc studies as archaeological reports and articles; structural,
architectural, material degradation analyses.
• Theoretical building rules, orders and modules of the architecture
of antiquity described by Vitruvius in De architectura libri decem.
3.4. Production workﬂow
This segment involved a team composed of different professionals
including computer scientists, graphic designers, surveyors,
modelers, archaeologists, architects and historians. The aim was
to study the ancient artifact with a multidisciplinary approach
and obtain its reconstruction through a range of proven activities
[17,20]. The production was divided in 2 steps, interpretation and
3D modelling, both performed almost simultaneously.
D. Ferdani et al. / Journal of Cultural Heritage 43 (2020) 129–143 133
Fig. 2. Digitization of the archaeological site of the Forum of Augustus and fragments of the marble decoration performed with IBM method.
3.4.1. Interpretation
All the sources acquired and organised in the previous phase
were, at this stage, analyzed and discussed together with the scientiﬁc
consultants, expert in the various disciplines (archaeology,
history, architecture, structures, etc.), in order to sketch a plausible
reconstructive hypothesis.
As a support tool for discussion we used the 3D models obtained
from the surveys in order to carry out graphic simulations and
extract accurate metric data. In this way it was possible to virtually
restore the marble fragments and hypothesize the original location
of the fragments through anastylosis processes. As in the case of the
porticoes or the attic of the Forum (the architectural element above
the colonnade embellished with decorative Caryatids alternating
with heads of Juppiter), they were partially digitally recomposed
by juxtaposing the scanned fragments of the decoration. The missing
parts were then deduced from rhythm and proportions of the
pieces, their size and Vitruvius’ architectural rules. It was necessary
to combine a large number of sources in order to formulate
the reconstructive hypothesis. Furthermore, for each architectural
element, part of the building, we had to reconstruct different physical
properties: shape, volume, size, materials, decorative apparatus
or architectural style.
Once the reconstructive hypothesis was formulated, the modelers
then have sketched the ﬁrst draft reconstruction of the assets.
This ﬁrst draft was then used for further investigation and veriﬁcation
before modeling the ﬁnal assets (see Section 3.4.2). The
continuous involvement and exchange between the different professionals
ensured the quality of the virtual reconstruction. Each
new version of the virtual reconstruction was veriﬁed until the
approval of the ﬁnal version discarding those that, during the work,
resulted less probable.
Fig. 3 illustrates part of the sources used to elaborate the interpretative
hypothesis of the temple and their relation with the
virtual reconstruction: D01-06: images documenting the main
architectural elements still present; D.07-09: architectural fragments
preserved within the Museum of the Imperial Forums; D.09:
Relevant fragment from the Ara Pietatis depicting the facade of the
temple of Mars Ultor in the Forum of Augustus; D.10: Table by the
architect G. B. Piranesi, depicting the Forum of Augustus at the end
of the 17th century; D.11-12: scientiﬁc studies from journals and
books. In order to set-up a reliable and “transparent” hypothesis, it
is necessary to track the whole process of the virtual reconstruction.
This means that not only the sources, but also the logical and deductive
processes of analysis and synthesis, that led us to elaborate the
virtual reconstruction from the interpretation of the sources, must
be mapped and recorded in a formal language.
Reliability of the reconstruction and transparency of the data
are primary requirements according to the principles of the London
Charter [21] and can be managed with different approaches
as long as the result is adequate. In our case we have adopted the
Extended Matrix,14 a visual node-based formal language that, being
based on a stratigraphic approach in use in archaeology, allowed the
archaeological documentation and the virtual reconstruction to be
connected [18,19]. In this way, for each property of each object, it
was possible to record the interpretative process used according to
the available sources (deduction on the basis of comparisons, parallelisms,
testimonies, etc.) and the reliability of the sources (see
Fig. 3).
3.4.2. 3D modelling
This activity consisted in the modelling of the 3D reconstructive
asset through the use of computer graphics and takes place
almost simultaneously with the interpretation and validation activity.
The ﬁnal output of the work will inevitably inﬂuence the
modelling approach (depending on target platform used to deploy
the interactive game or application). It usually affects the overall
workﬂow carried out on geometries, texturing, etc. when targeting
speciﬁc real-time output, also including aesthetic aspects. In fact,
virtual 3D environments created for standard or non-stereoscopic
visualization, generally require vast re-adaption for immersive VR
experiences targeting Digital Heritage ﬁeld. Designing and crafting
immersive virtual environments for such experiences may present
bottlenecks and huge efforts in terms of 3D modelling and/or optimization
tasks. Depending on locomotion models employed and
spatial constraints, speciﬁc elements of the 3D scene may capture
users’ attention or require higher complexity. The amount
of detail required to guarantee a consistent experience through
Head Mounted Display (HMD) can be overwhelming for large
3D reconstructed environments, especially when small teams are
involved in the workﬂow. For this reason, before starting the actual
3D modelling, a spatial prioritization map is produced in order
to optimize and drive the overall 3D assets workﬂow. In order
to employ such step, a preliminary 3D draft model of the 3D
14
http://www.osiris.itabc.cnr.it/extendedmatrix/osiris.itabc.cnr.it/extendedmatrix/.
134 D. Ferdani et al. / Journal of Cultural Heritage 43 (2020) 129–143
Fig. 3. Top row: visual mapping of the sources that allows to keep track of the reconstructive process making it transparent and scientiﬁcally accurate. At the top right: 3D
model of the temple of Mars Ultor in its current state of preservation. Medium row: some of the sources used for the reconstruction. (credits: images D.07-09: Mercati di
Traiano – Museo dei Fori Imperiali; images D.06 and D.10: Archive of the Sovrintendenza Capitolina). Bottom row: source-based reconstructive hypothesis of a column of
the temple and tracking of the sources using the extended matrix approach. For simplicity we have used, in this example, a synthetic version of the EM. node system.
environment is required. In our case we exploited the Forum of
Augustus – already reconstructed in the Trajan age 15 (see Section
3.1). Even if there are signiﬁcant differences between the
two models, the main architectural spaces and volumes were
15
http://keys2rome.eu/.
pretty similar and the quality of model sufﬁcient for our pur-
poses.
3.4.2.1. Spatial prioritization. Previous research on stereoscopic
visualization and depth perception in HMDs [36,10] combined with
the adopted locomotion model for the interactive VR experience
– can be exploited to perform a preliminary step to rank the vir-
D. Ferdani et al. / Journal of Cultural Heritage 43 (2020) 129–143 135
tual environment in terms of detail requirements. The main goal is
to minimize 3D modelling effort targeting immersive VR by prioritizing
the workﬂow for selected elements/areas of 3D scene,
depending on planned user locomotion. In order to produce a spatial
prioritization map, we took into account two target locomotion
models for immersive VR exploration: (a) teleport and (b) nodebased
locomotion [8,6].
Fig. 4 shows the overall prioritization map resulting from the
considered locomotion models, to support and guide 3D modelling
workﬂow. We deﬁned three different priority levels depending on
users’ potential proximity (d) during exploration: d <15 m (red); d
<30 m (green); d over 30 m (white). Regarding the Augustus Forum
3D reconstruction for instance, this guided modelers to focus their
effort on ﬂoor details (A) instead of roofs (B – e.g. porticoes). Thanks
to the spatial ranking, some elements like the columns of the main
entrance of Temple of Mars (2) resulted in light requirements in
terms of geometric detail, especially regarding capitals (see Fig. 4).
Same considerations apply to Victory statues (1 – top of the temple),
requiring minimal detail on both geometry and texturing level,
being very far from all possible user locations. Such an approach
resulted in a time-saving solution for 3D modelers, offering a small
team a valid support to prioritize the effort during the creation of
static elements of the 3D virtual environment.
Additional data can be employed to support or ﬁne-tune the
overall prioritization map: when available for instance, real users
propensities and spatial behaviors can be exploited. An immersive
VR application (“ovrWalker”16) was previously deployed during
public exhibits to explore a set of scenes – including the 3D reconstruction
of the Forum in the Trajan age (see Section 3.1). The
interactive application was enriched with a special software component
to anonymously capture casual user sessions.
During the public event TourismA 201817 for instance, the component
was able to collect large amounts of locomotion data, with
0.1 s temporal granularity. Since the adopted locomotion model in
ovrWalker offered visitors a completely free exploration within
the forum, the dataset was analysed and encoded to understand
users’ spatial behaviors [24,26,25] providing 3D modelers additional
information to reﬁne the modelling workﬂow and priorities.
3.4.2.2. Asset creation. Reconstructive 3D assets were modelled
and completed after the ﬁnal validation and the availability of the
digital replica of the site and artifacts (Section 3.3.1) facilitated the
process and ensured greater accuracy. The digital replica of the site
was fundamental in the design of the volumes according to the
correct ratios and measurements and in the positioning of the architectural
elements characterized by roman rhythms and proportions
(columns, architraves, steps, etc.).
The juxtaposition of the two models allowed to reconstruct the
geometries with great precision avoiding the typical problems that
can be generally found using two-dimensional drawings: inconsistency
between sections or plans, lack of information in the parts
not that are not represented, etc. (see Fig. 5).
In computer graphics, the denser the polygonal resolution of
the mesh, the greater the detail is. However, the topology and
the amount of geometry of the model strongly affects the performance
of the Graphics Processing Unit (GPU) and therefore it
is necessary to overcome the problem by using some optimization
techniques in order to minimise the workload and increase
the efﬁciency of real-time rendering. These optimizations must not
affect the aesthetic aspect but must always allow the recognition
of the represented artifact and its details, thus ensuring ﬁdelity and
realism of the model. Therefore, in order to achieve a compromise
16
http://osiris.itabc.cnr.it/scenebaker/index.php/projects/osglab/ovrwalker/.
17
http://www.tourisma.it/home-2/.
between smooth performance and realistic visuals, the optimization
can be performed borrowing different well-known techniques
from computer game design. In the reconstruction of the Forum
two main techniques were used to maintain a consistent detail: (1)
LODs18 for several elements and (2) normal mapping or parallax
occlusion mapping (POM) techniques to simulate additional detail
on surfaces (relief patterns, etc.).
In the ﬁrst case, different versions of the model are created
with a progressively lower polygonal resolution, which will then
be dynamically managed by the game’s graphic engine depending
on camera range [45]. In our case, the polygonal detail has been also
evaluated considering the spatial priority described in the previous
paragraph. In other words, depending on the rank, the various elements
of the virtual environment were modeled accordingly. For
example, ceilings, roofs or decorations on the top (11 m far from
the user’s point of view), have very simpliﬁed geometries. Normal
mapping and POM techniques are both used to simulate additional
3D detail through textures without increasing the geometrical complexity
and maintaining simpliﬁed and low-poly meshes.
Normal mapping is the best compromise between visual quality
and real-time performance. Normal map is a RGB texture which
simulates the complexity of the relief of the surfaces without having
to describe them topologically (e.g. bas-relief, mouldings, etc.). Each
texel in fact encodes normal information that can be exploited by
lighting model to simulate relief.
Parallax Occlusion Mapping or POM, is a similar but more
advanced technique which relies on a gray scale texture representing
a displacement map (or height map). It is an enhancement of
parallax mapping, used to procedurally create a 3D deﬁnition for
textured surfaces consistent with current eye position. This technique
is more computationally expensive on the GPU compared to
normal mapping, but offers a more consistent perception of relief
and surface micro-variations due to the parallax effect, when performed
on each eye of the HMD. We employed this technique only
on low-poly assets presenting complex mirco-variations very close
to user in terms of locomotion, like column shafts and the marble
slabs of the Forum ﬂoor.
In our case study all the normal and height maps produced internally
have been created through baking processes between models
with different LODs. In the case of the temple of Mars Ultor, the
high-resolution reconstructive model has a high geometric detail.
On the outer wall of the Cella, the ashlar decoration and the continuous
band with meander waves are modelled respecting the original
measurements recorded in the acquisition phase. The optimized
model used in the game instead has a more simpliﬁed geometry
where the details of the above elements is purely visual and determined
by normal map. The aesthetic perception is almost identical
and the formal coherence to the historical artifact is guaranteed
(see Fig. 6A).
Whenever possible, 3D models of very peculiar and complex
architectural elements such as the Caryatids of the atticus or the
Corinthian capitals of the Cella decorated with Pegasus protomes,
were produced from 3D survey. The resulting high-resolution models
have been simpliﬁed through polygonal decimation and then
optimized through semi automatic re-topology operations. This
approach allowed to obtain simpliﬁed meshes, consisting of quadrangular
faces offering better manipulation in CG software than
original triangulated high-poly models [39]. Furthermore during
this operation it is possible to decide the target number of polygons
that the model should have. Once regular and low-poly meshes
were obtained, several baking operations were performed to trans-
18
Level of detail.
136 D. Ferdani et al. / Journal of Cultural Heritage 43 (2020) 129–143
Fig. 4. Workﬂow prioritization map (left). Most visited locations of the previous 3D reconstruction of Forum of Augustus, resulting from locomotion data recorded during
public exhibits (right).
Fig. 5. 3D modelling of the Forum of Augustus. The creation of the reconstructive geometries is made directly on the reality-based model obtained with IBM techniques.
fer normal maps and to achieve the same visual result as the original
models (see Fig. 6B).
Sometimes additional manual modelling work was necessary in
order to virtually restore some missing parts or close small holes in
the mesh caused by under-sampling in the 3D survey. Such a result,
of high quality and great detail, allowed modelling team us to save a
lot of time. Indeed, modelling such complex geometry and textures
would have required laborious sculpting operations if carried out
manually.
3.4.3. Texturing
Recent game engines like Unreal Engine, Unity 3D, etc. take
advantage of a combination of textures to reproduce the physical
behaviour of surfaces in relation to the ﬂow of light. This texture
are called PBR or physically-based rendering. Among the most used
properties, we employed a metalness workﬂow (albedo, roughness,
metalness and emission) along with normal-maps, whose use is by
now part of modern game engines, in order to improve the appearance
and photorealism of interactive virtual 3D scenes. In order to
obtain photorealistic 3D models, different PBR textures type have
been created:
• Albedo represents the base chromatic information and it is
created either from photographs or through drawings made
by graphics (especially for lost decorations). The simulation
of patterns and colors of speciﬁc elements, such as frescoes,
stuccoes, marbles, etc. required comparison with the historicalarchaeological
documentation.
• Roughness determines the micro-relief that inﬂuences the
behaviour of light on the surface according to the type of material
represented (marble, bronze, travertine, etc.).
D. Ferdani et al. / Journal of Cultural Heritage 43 (2020) 129–143 137
Fig. 6. Optimization workﬂow: (A) example of texture used on low-poly 3D model for simulating high-poly surface and resulting rendering. The ashlar decoration on the
wall is created by normal maps while the vertical ﬂuting on Corinthian order columns is generated by height maps; (B) example of the image-based models optimization
workﬂow adopted to obtain realistic models from scanned museum objects.
• Metalness is used to set a material as metallic or non-metallic
and is used in combination with the roughness map to control
the amount of reﬂection according to speciﬁc material or surface
degradation due to consumption (braziers, candelabra, fences,
etc.).
• Emission is used to simulate a self-illumination of a model surface
by increasing the pixel intensity. We applied this map to the
glasses of the roman lantern in order to simulate the oil light and
portions of the Suburra.
The combination of these properties (as textures) determines
the surface behaviour of the model and how it will react to lighting.
These are applied to the model by means of a UV mapping
technique in which the polygons of the mesh are unwrapped and
associated with the two-dimensional plane of the texture. For the
creation of textures, some free online libraries were used to reproduce
the most common materials such as metals or marbles, but in
most cases it was necessary to create textures ad hoc through photographic
campaigns or performing digital painting as in the case
of the painted marble slabs.
In order to simulate some surfaces, such as the large wall in ashlar
of lapis Gabinum, that separated the Forum from the Suburra, 3D
survey was used. IBM techniques made it possible to obtain a highresolution
3D model of a well-preserved portion of the wall. Then,
in post production, it was possible to transfer the color and geometric
detail from the model to the textures of albedo, normal and
roughness using “backing procedures”. Finally, further corrections
in 2D editing software were necessary to make the resulting textures
“tileable”. This process was time consuming but allowed to
build a highly accurate and realistic result during real-time rendering
(see Fig. 7).
Creating textures for the whole Forum was laborious and complex.
The amount of complex mouldings and architectural elements
and the multitude of materials used, such as the different types
of white and coloured marble (white Carrara, pavonazzetto, giallo
antico, africano and cipollino) often combined in polycrome geometric
decorations (ﬂoors, walls of the Cella and Hall of Colossus),
required the production of several albedo textures arranged into
approximately 50 atlas (see Fig. 7).
Table 1
Relation among material evidences and their reconstructions.
Material evidence Reconstruction
Extant structure: archaeological evidence still
present in-situ
Structural virtual
reconstruction based
on archaeological
evidence in-situ
Fragmental virtual anastylosis: recomposition
of collapsed architectural fragments which
were found out of context and now preserved
in museums
Fragmental virtual
reconstruction.
Completion of
repositioned fragments
Lacking of physical evidence. Existence is
deduced either by a various necessary
conditions (roman styles, modules or
architectural rules) or from testimonies (i.e.
Ara Pietatis depicting the lost facade of the
temple of Mars Ultor)
Non-structural virtual
reconstruction based
on comparisons of
sources interpretation
3.4.4. Mapping the uncertainty
The 3D modelling is an integral part of the work of interpretation
and allows archaeologists to reﬁne the reconstructive hypotheses
by verifying in the reconstruction – which is nothing more than a
reverse engineering process (reconstruct the original project from
remains) – the effectiveness of the reconstructive hypotheses. The
work of interpretation, mapping of sources and their organization
in a formal language can be represented on the 3D model through
the use of proxies (see Fig. 8) that immediately show us the different
levels of reliability of our reconstruction making its historical background
transparent and allowing us to distinguish certain from the
uncertain, the necessary condition from interpretations [53,20]).
The levels of reliability were established as in Table 1 at page 12.
This uncertainty mapping is a working tool that allows the scientiﬁc
community to evaluate and track the reconstruction work.
However, appropriate graphic representations could be implemented
into video games and could be interrogated by players in
the “museum mode”. This mode, being deprived of video games
dynamics, allows the user to freely explore the spaces and get information
about the history of the site. The integration of information
related to the interpretative study and the historical reliability of
the reconstruction (already developed for interactive applications
but not for games [19]) could be an added dissemination strategy
for future productions.
138 D. Ferdani et al. / Journal of Cultural Heritage 43 (2020) 129–143
Fig. 7. Texturing. Top row: workﬂow adopted for creating textures from 3D scanned models. Bottom row: Textured model of the forum. The section shows the complexity
of the architectural order and the polychrome marble decoration of the interiors and porticoes.
3.5. Level creation and authoring
These activities are carried out by the level designer inside the
game engine with the support of 3D modellers, historical consultants,
game designers and user experience experts. The level – or
“game level” – represents an arrangement of static and dynamic 3D
assets produced by the previous phases. This segment concerns the
communicative and aesthetic aspect in which the work of virtual
reconstruction is synthesized in a visual representation. In other
words, the reconstructive model is imported into the game engine
– software that allows to view real-time scenarios and game assets
and perform programming operations that will deﬁne the behavior
and dynamics of the game – to assemble the virtual reconstructive
scenography.
After assembling the environment, further reﬁnements are
needed to set the graphic rendering that is based on the creative
choices made during the design phase of the game: scene dressing,
real-time rendering and soundscape set-up.
3.5.1. Scene dressing
In this step, props – dynamic object that can be moved and or
interact with, are imported and instanced in the game scenario.
In the case of the Forum of Augustus several props, fundamental
to understand the function of architectural spaces and enrich
them with realistic details, were added. The big exedrae were ﬁlled
with roman furniture like wooden tribunals and curule seats while
the small one with cabinets for papyruses and wax tablets, table
D. Ferdani et al. / Journal of Cultural Heritage 43 (2020) 129–143 139
Fig. 8. Proxy model of the Forum of Augustus. The level of certainty are indicated by the colours which distinguish extant structures and repositioned fragments from their
completion and reconstruction. The green colour instead, maps all those parts for which we have no structural or archaeological evidence but their reconstruction is entrusted
to comparisons or interpreted sources.
and chairs. All the interiors were provided with braziers, oil lamps,
torches and candelabra and statues. Furthermore, a building site
area were deliberately inserted to show the south-west part of
the Forum still under construction. Among the assets, in addition
to those already mentioned, a roman crane, scaffolding, centering
arches and other wooden infrastructures used by roman in the
building site were modelled.
All these objects were made from scratch with the help of historical
documentation (in the case of the crane we used different
iconographic sources like the relief sculpted on the decoration
of the tomb of the Haterii family which is preserved in the Vatican
Museums). This expedient allowed us to address some issues
related to the technologies and practices adopted by the Romans.
3.5.2. Real-time rendering
Another relevant aspect concerns the real-time rendering,
highly dependent on the chosen game engine (e.g. Unreal Engine
4, Unity3D, etc.). In this phase the graphic style and realism of the
experience is realized by setting up materials and lighting:
• Creation of materials associated with each 3D model (marble,
stone, metal, glass, etc.) from PBR maps (see Section 3.4.3) to
simulate or approximate its physical behavior (reﬂection, transparency
or emission) inside the game engine.
• Lighting set-up by creating, positioning and tuning light sources
in the level.
In order to ensure the historical consistency, these activities are
often validated by specialists (e.g. historical consultants). Regarding
the interiors of the Forum of Augustus for instance, it was important
to choose which, among different types of light sources (braziers,
torches, lanterns, oil lamps) were the most suitable for speciﬁc locations
– see Fig. 9 – and their parameters to simulate the ancient light
performance (temperature or light color) according to real data (see
for instance [48]).
These tasks – including also previous scene dressing activities –
must also ensure robust performances for the immersive VR application/game
being created. In our case, we ﬁne-tuned the process
to maintain a range of 70–90 fps19 on consumer-level HMDs (e.g.
Oculus Rift) and hardware across the entire level of the Forum of
Augustus.
3.5.3. Soundscape
These activities allow to create a collection of sounds arranged in
speciﬁc locations of the level in order to enhance the sense of presence
[57,13] for the immersive experience. Speciﬁcally, ambient
sounds are background sounds with speciﬁc location and radius,
used to create a sense of atmosphere [58] (wind blowing, barking,
street noises, etc.). Sound sources – with given falloff distance –
can be also attached to dynamic props inside the game engine to
facilitate re-use in the level (e.g. braziers).
Once again, a validation process is required in order to maintain
consistency of the overall soundscape. For instance, ambient
sounds associated to the Suburra area (close to the guard-post) were
designed with the support of historical consultants: speciﬁcally,
during night time only sporadic sounds (e.g. barking) were probably
heard due to the area being disreputable (lower-class area)
and thus less crowded during that hours. In our case, the validation
process was carried out on sound types only, without considering
propagation properties of different materials (see for instance [63]).
4. Results and discussion
This section brieﬂy describes multiple resulting immersive VR
products, exploiting produced 3D assets, including applied VR
games and explorative applications.
4.1. A Night in the Forum
A Night in the Forum is an Educational Environmental Narrative
(EEN) Game for PlayStation® VR, set in the Rome of Augustan age,
whose goal is to make the player understand the complexity of the
administration of Imperial Rome. The game experience is based
on the “environmental storytelling” and the “learning-by-doing”
19
Frames-per-second.
140 D. Ferdani et al. / Journal of Cultural Heritage 43 (2020) 129–143
Fig. 9. Real-time rendering of the Forum of Augustus using different lighting conditions using Unreal Engine 4.
approaches [30,37]. The player is thrown into Forum of Augustus
where he has to spend an entire night exploring the reconstructed
ancient architectures and life spaces (see Fig. 10).
Furthermore he/she has to accomplish several tasks interacting
with digital reproductions of real cultural artifacts and carrying out
authentic assignments. The user plays the role of a guardian who
has to watch both the public and sacred spaces of the Forum at
night-time and arrange them for the next day’s opening (e.g. sort
out the tablets in the archives, light braziers in the Hall, etc.). During
the journey the player discovers information and stories about the
Forum and its elements (statues, decorations, ritual ceremonies,
etc.).
The game was developed by VRTRON20 and CNR ITABC (3D modelling
of historical assets) with the collaboration of the Museo dei
20
vrtron.com.
Fori Imperiali – Rome (historical consultant) within the European
Project REVEAL.21 and is available on the playstation store 22
4.2. Forum of Augustus VR
This immersive VR application is purely explorative, developed
using Unreal Engine 4.23 The game engine allowed to fully
exploit modern rendering capabilities, including high-ﬁdelity PBR
materials, spatialized soundscape and a few interactions with the
environment through VR controllers (e.g.: opening doors). The locomotion
model employed is point-and-teleport [8], providing a safe
exploration for a wide range of users. Virtual hands (VR controllers)
are included for teleport actions and provide basic interactions with
the environment. Starting from produced 3D assets, multiple loco-
21
revealvr.eu.
22
https://store.playstation.com.
23
https://www.unrealengine.com.
D. Ferdani et al. / Journal of Cultural Heritage 43 (2020) 129–143 141
Fig. 10. From top to bottom: (1) “A Night in the Forum” – Cover of the VR game and screenshot of the gameplay; (2) “Forum of Augustus VR” – Deﬁnition of basic colliders,
locomotion areas (green) inside the game editor and in-game captures while using HMD and VR controllers; (3) “ VR Forum” – Misplacing items inside the game editor and
in-game multi-scale manipulation to solve all items.
motion areas (see Fig. 10) were deﬁned inside the game editor
through the arrangement of multiple shapes/colliders. A few spatialized
sounds were also placed into the virtual scene to create
immersive soundscapes.
4.3. VR Forum
This immersive VR puzzle-game, presented for the ﬁrst time in
TourismA 2019, a fair on archaeology and cultural tourism held in
Florence,24 is based on the VR game model [29] that combines real
walking and multi-scale 6DOF gaming. The model25 offers users an
immersive, physical and engaging re-contextualization challenge
within architectural and archaeological contexts. The player has to
ﬁnd multiple misplaced items scattered across the Forum of Augustus
and put them back in their original locations by performing
basic manipulation tasks (grab and release – using VR controllers).
24
www.tourisma.it.
25
http://osiris.itabc.cnr.it/scenebaker/index.php/projects/uvr/.
142 D. Ferdani et al. / Journal of Cultural Heritage 43 (2020) 129–143
Once an item is solved (placed in its original position), the virtual
space changes scale: this offers the user a new spatial perception
of the virtual environment, following a mechanic inspired by the
novel “Alice in Wonderland”. The game model suits micro- and
macro-scale contexts, and minimizes motion sickness by removing
artiﬁcial locomotion, since each item is solved through physical
motions within the tracked area boundaries (for further details
see reference paper). For the exhibit, the game exploited the latest
3D reconstruction and assets of Augustus Forum and it was set
at sunset (see ﬁgure 18). The setup included an Oculus Rift CV1
and a physical tracked area of 2.5 × 2.5 m in order to deploy the
immersive VR game. A collection of items with different sizes was
arranged together with a few archaeologists using the game editor
for the re-contextualization game: the roof of the temple, a column
of the temple, the victory statue, the colossus of Augustus, the main
door of the temple and the statue of Mars and Venus.
4.4. Final remarks
The forum was a public space, generally crowded: some sort
of crowd simulation is eventually expected for the described VR
applications. Several techniques have emerged in the last decade to
populate a virtual space, including those dealing with immersive VR
products (see [50,22]). For instance in [28] a micro-dynamics model
was adopted to populate the reconstruction of ancient Bononia
and to support the gameplay. We did not include such kind of
simulation in our applications for many reasons: philological character
modeling targeting immersive VR can be time-consuming
([9]); crowd micro-dynamics are computationally intensive (see
for instance [23]); according to historical sources, the access to
the forum was forbidden during the times our applications/games
are set. Regarding user evaluation, we carried out a preliminary
user testing on VR Forum application ([29]) with high school students
and during public events (“TourismA 2019”), in order to study
effectiveness of the multi-scale puzzle approach in VR, obtaining
encouraging results. Regarding “A night in the Forum” we planned
a series of evaluation tests at the Imperial Fora Museum once
the applied VR game will be on display (2020), including learning
assessment.
5. Conclusions
Virtual reconstruction is a great didactic tool as it improves
cognitive processes making the historical and archaeological
data easily comprehensible to anyone and within a video game
this potential is reinforced by the dynamics of storytelling,
and learning-by-doing. However, as emerges from the “picture”
described above, the virtual reconstruction of the past imposes
many limitations and great effort to ensure the consistency and
reliability of the reconstructive hypothesis. The latter indeed prolong
the production times, increase costs, require the involvement
of experts in various cultural ﬁelds and cross-domain professionals,
capable of dealing with both computer graphics as well as archaeology
worlds.
Another limitation for this class of cultural applications is the
creation of the 3D virtual environment. The design of historical
architecture and user navigation spaces impose some constraints:
while in video games design there is no limit to imagination and
script and scenarios are created according to the game needs and
logic, in the serious games segment this is not possible. On the contrary,
it’s the game that has to suit and adapt to the historically
validated 3D environment, to maintain consistency.
The research can play a decisive role in helping developers – in
particular those without the large funding available to big game
entertainment companies – through the creation of collections of
cultural assets or supporting them during the application/game
development. Furthermore, this can fuel research activities targeting
immersive VR segment by providing fertile playground
for developers, game designers, researchers and many others, for
instance to assess novel interaction models, game mechanics or
content creation pipelines.
During last decade, 3D modeling and visualization technologies
did prove to enhance cognitive and interpretative processes for a
better understanding of archaeological contexts, helping scholars
to structure complex sources and test interpretative reasoning in
virtual reality models [35]. Indeed many universities and research
centers, during their activities, put their effort into the reconstruction
of sites and artifacts according to scientiﬁc approaches. The
creation of libraries that enable the availability of reconstructive
models of cultural assets – including linked metadata related to
both the characteristics of the object (period, function, cultural context,
etc.) and the sources used – could become useful for companies
or other researchers to boost up production workﬂows. We are not
only referring to architectural 3D assets, for which this approach
is intrinsically limited, especially to dynamic elements commonly
referred as props. Architectural 3D assets in fact, are usually unique
entities and their re-use can be considered only for some building
components (the ones that are typically serialized or obeying
speciﬁc architectural rules – e.g. capitals or entablatures). In the
case of the Forum of Augustus, many props were crafted – tables,
chairs, lanterns, braziers, statues, columns – following a philological
process which could be re-used in future projects. This logic of
sharing, according to a policy to be deﬁned, could follow the ones
adopted for the creation of digital heritage collections. The latter
bring together contributions from all countries with the purpose of
conservation and dissemination thanks to European projects such
as Europeana,26 Carare,27 3D Icons28 and many others.
Acknowledgements
This article is the result of 3 year of work and research. The
authors want to thank all the colleagues of the REVEAL consortium,
especially the VTRON company and the Mercati di Traiano –
Museo dei Fori Imperiali for the fruitful collaboration and the fun
experience of a “Night in the Forum”.
Appendix A. Supplementary data
Supplementary data associated with this article can be found, in
the online version, at https://doi.org/10.1016/j.culher.2019.12.004.
References
[1] E.F. Anderson, L. McLoughlin, F. Liarokapis, C. Peters, P. Petridis, S. De Freitas,
Developing serious games for cultural heritage: a state-of-the-art review,
Virtual Real. 14 (2010) 255–275.
[2] J. Barcelò, M. Forte, D. Sanders, Virtual reality in archaeology; british archaeological
reports, Int. Ser. 843 (2000).
[3] M.K. Bekele, R. Pierdicca, E. Frontoni, E.S. Malinverni, J. Gain, A survey of augmented,
virtual, and mixed reality for cultural heritage, J. Comput. Cult. Herit.
(JOCCH) 11 (2018) 7.
[4] F. Bellotti, R. Berta, A. De Gloria, A. D’ursi, V. Fiore, A serious game model for
cultural heritage, J. Comput. Cult. Herit. (JOCCH) 5 (2012) 17.
[5] F. Bellotti, R. Berta, A. De Gloria, G. Panizza, M. Pellegrino, L. Primavera, Designing
serious games for cultural heritage purposes, Digit. Games Learn. (2011)
252–276.
[6] C. Boletsis, The new era of virtual reality locomotion: a systematic literature
review of techniques and a proposed typology, Multimodal Technol. Interact.
1 (2017) 24.
26
https://www.europeana.eu.
27
https://www.carare.eu/.
28
http://3dicons-project.eu/.
D. Ferdani et al. / Journal of Cultural Heritage 43 (2020) 129–143 143
[7] F. Borgogni, A. Ippolito, I modelli 3d nei rilievi di architettura. Metodologie
integrate per il rilievo, il disegno, la modellazione dell’architettura e della città,
Gangemi Editore, Roma, 2011.
[8] E. Bozgeyikli, A. Raij, S. Katkoori, R. Dubey, Point & teleport locomotion technique
for virtual reality, Proceedings of the 2016 Annual Symposium on
Computer–Human Interaction in Play (2016) 205–216.
[9] G. Bozzelli, A. Raia, S. Ricciardi, M. De Nino, N. Barile, M. Perrella, M. Tramontano,
A. Pagano, A. Palombini, An integrated vr/ar framework for user-centric
interactive experience of cultural heritage: the arkaevision project, in: Digital
Applications in Archaeology and Cultural Heritage, 2019.
[10] L.E. Buck, M.K. Young, B. Bodenheimer, A comparison of distance
estimation in hmd-based virtual environments with different hmdbased
conditions, ACM Trans. Appl. Percept. 15 (2018) 21:1–21:15,
http://dx.doi.org/10.1145/3196885.
[11] M. Carrozzino, M. Bergamasco, Beyond virtual museums: experiencing immersive
virtual reality in real museums, J. Cult. Herit. 11 (2010) 452–458.
[12] E. Champion, Entertaining the similarities and distinctions between serious
games and virtual heritage projects, Entertain. Comput. 14 (2016) 67–74.
[13] T. Chandrasekera, S.Y. Yoon, N. D’Souza, Virtual environments with soundscapes:
a study on immersion and effects of spatial abilities, Environ. Plan. B:
Plan. Des. 42 (2015) 1003–1019.
[14] L. Charter, The London Charter for the Computer-Based Visualisation of Cultural
Heritage, 2009, Online at: www.londoncharter.org.
[15] D. Christopoulos, P. Mavridis, A. Andreadis, J.N. Karigiannis, Using virtual
environments to tell the story: the battle of thermopylae, in: 2011 Third International
Conference on Games and Virtual Worlds for Serious Applications,
IEEE, 2011, pp. 84–91.
[16] L.T. De Paolis, L. Oliva, P. Vecchio, et al., Otranto in the middle ages: a serious
game for the edutainment, Int. J. Inf. Educ. Technol. 1 (2011) 47.
[17] N. Dell’Unto, A. Leander, M. Dellepiane, M. Callieri, D. Ferdani, S. Lindgren, Digital
reconstruction and visualization in archaeology: case-study drawn from the
work of the Swedish pompeii project, in: 2013 Digital Heritage International
Congress (DigitalHeritage), IEEE, 2013, pp. 621–628.
[18] E. Demetrescu, Virtual reconstruction as a scientiﬁc tool, in: Digital Research
and Education in Architectural Heritage, Springer, 2018, pp. 102–116.
[19] E. Demetrescu, B. Fanini, A white-box framework to oversee archaeological
virtual reconstructions in space and time: methods and tools, J. Archaeol. Sci.:
Rep. 14 (2017) 500–514.
[20] E. Demetrescu, D. Ferdani, N. Dell’Unto, A.M.L. Touati, S. Lindgren, Reconstructing
the original splendour of the house of caecilius iucundus. A complete
methodology for virtual archaeology aimed at digital exhibition, SCIRES-IT-Sci.
Res. Inf. Technol. 6 (2016) 51–66.
[21] H. Denard, A new introduction to the london charter, in: Paradata and Transparency
in Virtual Heritage, Routledge, 2016, pp. 83–98.
[22] P. Dickinson, K. Gerling, K. Hicks, J. Murray, J. Shearer, J. Greenwood, Virtual
reality crowd simulation: effects of agent density on user experience and
behaviour, Virtual Real. 23 (2019) 19–32.
[23] B. Fanini, L. Calori, 3d interactive visualization of crowd simulations at urban
scale, in: 9◦
Congresso Città e Territorio Virtuale, Roma, 2, 3 e 4 ottobre 2013,
Università degli Studi Roma Tre, 2014, pp. 276–284.
[24] B. Fanini, L. Cinque, Encoding vr sessions: image-based techniques to
record and inspect immersive experiences, in: 3rd Digital Heritage
International Congress (DH2018) Held Jointly With 24th International Conference
on Virtual Systems and Multimedia (VSMM 2018), IEEE, 2018,
http://dx.doi.org/10.1109/DigitalHeritage.2018.8810070.
[25] B. Fanini, L. Cinque, Encoding immersive sessions for online, interactive vr analytics,
Virtual Real. (2019), http://dx.doi.org/10.1007/s10055-019-00405-w.
[26] B. Fanini, L. Cinque, An image-based encoding to record and track immersive
vr sessions, in: S. Misra, O. Gervasi, B. Murgante, E. Stankova, V. Korkhov, C.
Torre, A.M.A. Rocha, D. Taniar, B.O. Apduhan, E. Tarantino (Eds.), Computational
Science and Its Applications – ICCSA 2019, Springer International Publishing,
2019, pp. 299–310.
[27] B. Fanini, E. d’Annibale, E. Demetrescu, D. Ferdani, A. Pagano, Engaging and
shared gesture-based interaction for museums the case study of k2r international
expo in Rome, in: 2015 Digital Heritage, IEEE, 2015, pp. 263–270.
[28] B. Fanini, A. Pagano, Interface design for serious game visual strategies the case
study of “imago bononiae”, in: 2015 Digital Heritage, IEEE, 2015, pp. 623–626.
[29] B. Fanini, A. Pagano, D. Ferdani, A novel immersive vr game model for recontextualization
in virtual environments: the vr model, Multimodal Technol.
Interact. (2018), http://dx.doi.org/10.3390/mti2020020.
[30] C. Fernández-Vara, Game Spaces Speak Volumes: Indexical Storytelling, 2011.
[31] M. Forlani, E. Pietroni, L. Rescic, C. Rufa, F. Antinucci, Game embrace siegecraft
in “protect the walls!”, SCIRES-IT-Sci. Res. Inf. Technol. 6 (2016) 31–40.
[32] M. Forte, S. Pescarin, E. Pietroni, C. Rufa, Multiuser virtual reality museum application
for cultural heritage: the ancient via ﬂaminia project, Proceedings of The
8th International Symposium on Virtual Reality (2007).
[33] F. Gabellone, The reconstruction of archaeological contexts: a dialectical relationship
between historical-aesthetic values and principles of architecture, in:
Sensing the Past, Springer, 2017, pp. 395–417.
[34] F. Gabellone, A. Lanorte, N. Masini, R. Lasaponara, From remote sensing to a serious
game: digital reconstruction of an abandoned medieval village in southern
italy, J. Cult. Herit. 23 (2017) 63–70.
[35] S. Hermon, Scientiﬁc method, chaîne opératoire and visualization: 3d modelling
as a research tool in archaeology, in: Paradata and Transparency in Virtual
Heritage, Routledge, 2016, pp. 39–48.
[36] J.A. Jones, I.I. Swan, J.E. Singh, G. Kolstad, E.S.R. Ellis, The effects of virtual
reality, augmented reality, and motion parallax on egocentric depth
perception, in: Proceedings of the 5th Symposium on Applied Perception
in Graphics and Visualization, ACM, New York, NY, USA, 2008, pp. 9–14,
http://dx.doi.org/10.1145/1394281.1394283.
[37] M. Kebritchi, et al., Examining the pedagogical foundations of modern educational
computer games, Comput. Educ. 51 (2008) 1729–1743.
[38] G. Kontogianni, A. Georgopoulos, Exploiting textured 3d models for developing
serious games, Int. Arch. Photogramm. Remote Sens. Spat. Inf. Sci. 40 (2015).
[39] G. Kontogianni, C. Koutsaftis, M. Skamantzari, C. Chrysanthopoulou, A. Georgopoulos,
Utilising 3d realistic models in serious games for cultural heritage,
Int. J. Comput. Methods Herit. Sci. (IJCMHS) 1 (2017) 21–46.
[40] C. Kwon, Veriﬁcation of the possibility and effectiveness of experiential learning
using hmd-based immersive vr technologies, Virtual Real. 23 (2019) 101–118.
[41] E. La Rocca, La nuova immagine dei fori imperiali, Römische Mitteilungen 108
(2001) 171–213.
[42] N. Lercari, L. Onsurez, J. Schultz, et al., Multimodal reconstruction of landscape
in serious games for heritage: an insight on the creation of fort ross virtual
warehouse serious game, in: 2013 Digital Heritage International Congress (DigitalHeritage),
IEEE, 2013, pp. 231–238.
[43] F. Liarokapis, P. Kouˇril, P. Agraﬁotis, S. Demesticha, J. Chmelik, D. Skarlatos,
3d modelling and mapping for virtual exploration of underwater archaeology
assets, Int. Arch. Photogramm. Remote Sens. Spat. Inf. Sci. 42 (2017) 425.
[44] V.M. Lopez-Menchero, A. Grande, The principles of the seville charter, CIPA
Symposium Proceedings (2011) 2–6.
[45] D. Luebke, M. Reddy, J.D. Cohen, A. Varshney, B. Watson, R. Huebner, Level of
Detail for 3D Graphics, Morgan Kaufmann, 2003.
[46] J. Maïm, S. Haegler, B. Yersin, P. Mueller, D. Thalmann, L. Van Gool, Populating
ancient pompeii with crowds of virtual Romans, Proceedings of the 8th International
Symposium on Virtual Reality, Archeology and Cultural Heritage-VAST
(2007).
[47] M. Mortara, C.E. Catalano, F. Bellotti, G. Fiucci, M. Houry-Panchetti, P. Petridis,
Learning cultural heritage by serious games, J. Cult. Herit. 15 (2014) 318–325.
[48] D. Moullou, L.T. Doulos, F.V. Topalis, Artiﬁcial light sources in roman, byzantine,
and post-byzantine eras: an evaluation of their performance, Chronos 32 (2015)
119–132.
[49] J. Parong, R.E. Mayer, Learning science in immersive virtual reality, J. Educ.
Psychol. 110 (2018) 785.
[50] N. Pelechano, J.M. Allbecky, Feeling crowded yet?: crowd simulations for vr,
2016 IEEE Virtual Humans and Crowds for Immersive Environments (VHCIE)
(2016) 17–21.
[51] S. Pescarin, Reconstructing Ancient Landscape, Archaeolingua Budapest, 2009.
[52] S. Pescarin, Museums and virtual museums in Europe: reaching expectations,
SCIRES-IT-Sci. Res. Inf. Technol. 4 (2014) 131–140.
[53] S. Pescarin, D. Ferdani, B. Fanini, I. Cerato, G.L. Baldassari, Handling transparency
in 3d reconstructed on line environments: aquae patavinae vr case study, CAA
2012 (2011).
[54] E. Pietroni, A. Pagano, B. Fanini, Ux designer and software developer at the
mirror: assessing sensory immersion and emotional involvement in virtual
museums, Stud. Digit. Herit. 2 (2018) 13–41.
[55] P. Reilly, Towards a virtual archaeology, in: Computer Applications in Archaeology,
British Archaeological Reports, Oxford, 1990, pp. 133–139.
[56] F. Remondino, S. Campana, 3d recording and modelling in archaeology and
cultural heritage, BAR Int. Ser. 2598 (2014) 111–127.
[57] G. Seraﬁn, S. Seraﬁn, Sound Design to Enhance Presence in Photorealistic Virtual
Reality, Georgia Institute of Technology, 2004.
[58] D. Sonnenschein, Sound Design: The Expressive Power of Music, Voice, and
Sound Effects in Cinema, Michael Wiese Productions, 2001.
[59] C. Sternig, M. Spitzer, M. Ebner, Learning in a virtual environment: implementation
and evaluation of a vr math-game, in: Virtual and Augmented
Reality: Concepts, Methodologies, Tools, and Applications, IGI Global, 2018,
pp. 1288–1312.
[60] L. Ungaro, Storia, mito, rappresentazione: il programma ﬁgurativo del foro
di augusto e l’aula del colosso. Storia, mito, rappresentazione: il programma
ﬁgurativo del Foro di Augusto e l’Aula del Colosso, 2008, pp. 399–417.
[61] L. Ungaro, M. Milella, M. Vitti, Il sistema museale dei fori imperiali ei mercati
di traiano. Simulacra Romae, Roma y las capitales provinciales del Occidente
Europeo. Estudios Arqueológicos, 2004, pp. 11–47.
[62] F. Valls, E. Redondo, D. Fonseca, P. Garcia-Almirall, J. Subirós, Videogame
technology in architecture education, International Conference on
Human–Computer Interaction (2016) 436–447.
[63] J. Veitch, Soundscape of the street: architectural acoustics in ostia, in: Senses
of the Empire, Routledge, 2017, pp. 54–70.
[64] A. Vourvopoulos, F. Liarokapis, P. Petridis, Brain-controlled serious games for
cultural heritage, in: 2012 18th International Conference on Virtual Systems
and Multimedia, IEEE, 2012, pp. 291–298.