Hostname: page-component-89b8bd64d-shngb Total loading time: 0 Render date: 2026-05-07T13:18:43.446Z Has data issue: false hasContentIssue false
  • English
  • Français

Re-viewing Pompeian domestic space through combined virtual reality-based eye tracking and 3D GIS

Published online by Cambridge University Press:  24 March 2022

Danilo M. Campanaro Open the ORCID record for Danilo M. Campanaro [Opens in a new window]  and
Giacomo Landeschi
Danilo M. Campanaro*
Affiliation:
Department of Archaeology and Ancient History, Lund University, Sweden
Giacomo Landeschi
Affiliation:
Department of Archaeology and Ancient History, Lund University, Sweden
*
*Author for correspondence ✉ danilo_marco.campanaro@klass.lu.se
Rights & Permissions [Opens in a new window]

Abstract

This article presents an innovative methodology in which virtual reality-based eye-tracking techniques and 3D Geographical Information Systems are employed to record and measure human visual attention within the virtually reconstructed space of a Pompeian house.

Keywords

Pompeiieye tracking3D GISvirtual reality

Information

Type
Project Gallery
Information
Antiquity , Volume 96 , Issue 386 , April 2022 , pp. 479 - 486
Creative Commons
Creative Common License - CCCreative Common License - BYCreative Common License - NCCreative Common License - ND
This is an Open Access article, distributed under the terms of the Creative Commons Attribution-NonCommercial-NoDerivatives licence (https://creativecommons.org/licenses/by-nc-nd/4.0/), which permits non-commercial re-use, distribution, and reproduction in any medium, provided the original work is unaltered and is properly cited. The written permission of Cambridge University Press must be obtained for commercial re-use or in order to create a derivative work.
Copyright
Copyright © The Author(s), 2022. Published by Cambridge University Press on behalf of Antiquity Publications Ltd

Introduction and state of the art

Ancient literary sources (e.g. Cic. Att. 2.3.2; Cic. De Or. 1.39.179; Plin. Ep. 2.17; Plin. Ep. 5.6; Sen. Cont. 5.5; Stat. Silv. 1.3; Stat. Silv. 2.2; Vitr. 6.3.10) highlight the importance of the view and the act of viewing within Roman domestic architecture. Consequently, scholars have studied such sources in the context of the material remains of Pompeii and Herculaneum. The results suggest that optical axiality played a key role in the apperception of the Roman house and its socio-political significance (Drerup Reference Drerup1959; Bek Reference Bek1980; see also Hales Reference Hales2003: 107–22; Trentin Reference Trentin2019), ultimately envisaging a type of “vision mania” (Clarke Reference Clarke1991: 21). Questions such as which spaces within Roman houses were visible to visitors and how house design and visual space contributed to the social atmosphere, however, are still far from being fully answered.

Archaeologists have long been interested in visibility studies (e.g. Wheatley Reference Wheatley, Lock and Stancic1995; Llobera Reference Llobera1996). Initially divided between quantitative or experiential approaches, over the last 15 years, scholars have begun to combine different techniques, such as Geographical Information Systems (GIS), 3D modelling and virtual reality (e.g. Paliou Reference Paliou, Bevan and Lake2013; Landeschi et al. Reference Landeschi2016; Opitz Reference Opitz2017; Richards-Rissetto Reference Richards-Rissetto2017). Drawing on this tradition, the present study examines the possibility of integrating in 3D GIS data related to human visual attention (i.e. a series of cognitive operations that allow the selection of relevant information from a visual scene). These data, measured with an eye tracker, allow assessment of which parts of a complete reconstruction of a Pompeian house draw the attention of the user while moving in the virtual space.

Virtual reality-based eye tracking is an emerging technology. There are, so far, no published examples investigating the possibility of integrating the data generated by an eye-tracker embedded in a virtual reality headset into a GIS environment.

Methodology

We developed the work pipeline by integrating different products in order to: measure the visual attention of the user in a virtual reality environment; generate a data output for export into third-party software; and import the output data into a 3D GIS environment. The apparatus used in the trial was the Vive Pro Eye headset with integrated eye tracker.

A 3D reconstruction of the House of the Greek Epigrams in Pompeii (V 1,18) was built using 3D modelling software (3D Studio Max) (Campanaro Reference Campanaro2021) and successively imported into Unity, a cross-platform game engine for virtual reality experiences (Figures 1 & 2a). This software enabled a physically based lighting simulation. For the specific scope of this research, two very different scenarios were chosen: dawn on the winter solstice and noon on the summer solstice. The system recorded the visual experience of the participants in the trial (five age-diverse, Caucasian females). Output data, analytical insights and direct measurements were obtained using the software Cognitive3D (Figure 2b).

Figure 1.

The 3D model of the House of the Greek Epigrams, imported in Unity (figure by the authors).

Figure 2.

Workflow for the integration of eye-tracking data into a GIS environment (figure by the authors).

It is of paramount importance to geolocate the visual behaviour of a user by integrating the collected 3D data into a GIS environment. Specifically, this process may allow specialists to take advantage of the analytical power offered by this system (e.g. statistical analysis, 3D density maps, line-of-sight, distribution maps, multi-layered analysis) and to investigate data critically against other important layers of information (e.g. light metrics calculated using a lighting simulation tool [Figure 3], audio metrics, graffiti, household artefacts). Information derived from Cognitive3D was imported into a GIS environment (ArcGIS Pro) (Figure 2c) as JSON files—a standard human-readable exchange data format—comprising three datasets: gaze (accurate tracking of the head position of the user); fixation (detailed understanding of what users are focusing on when their eyes are relatively stable, where the stable point is called ‘fixation’); and event (duration, beginning and end of a single session) (Figure 4a). Each point in a dataset is identified by a triplet representing the (x, y, z) values according to the Unity coordinate system. Due to differences in the Unity and ArcGIS coordinate systems, transformation of the Unity coordinates to ArcGIS coordinates was required (Figure 4b). Once created, the datasets were imported as an ASCII file, along with the 3D model used for the simulation in Unity3D. The relative position of the eye tracker-measured points perfectly matched the space in which the users moved virtually while wearing the VR headset (Figures 5 & 6).

Figure 3.

Example of light metrics. Level of illumination (LUX) in the reconstructed house of the Greek Epigrams, calculated for four different hours of the winter solstice (unpublished data) (figure by the authors).

Figure 4.

JSON datasets exported from Cognitive3D, containing results of a single user's visual experience (a). Eye-tracking-derived datasets are edited with a text editor in order to adjust x,y,z column placements, according to the ArcGIS PRO combination scheme (b).

Figure 5.

Exported data are visualised in ArcGIS PRO as point shapefiles. The path of the user exploring the virtual environment of the reconstructed Pompeian house is displayed in the right spatial relation with the original building (figure by the authors).

Figure 6.

Gaze and fixation is shown with respect to the surrounding space. Max fixation radius is displayed with graduated symbols based on the size of the radius (figure by the authors).

Results and future prospects

In this trial, users’ visual experiences (gaze, fixation and movement) within a virtually reconstructed Pompeian house have been collected and imported into a GIS environment. Handling virtual reality eye tracker-derived data in a GIS space may present interesting new scenarios for research and analysis of human/environment interaction (e.g. to investigate how illumination influenced the social experience of elements such as graffiti and wall paintings). It is possible to query the geodatabase and automatically select areas with higher levels of visual attention, and to understand in which order objects are seen and the positions taken in the space by the user. Multiple human sessions can be integrated, and a large quantity of data can be managed in order to detect patterns that may relate to particular features of the examined virtual space. Furthermore, by managing these data in 3D GIS, an individual user's experience can be associated with a descriptive record entry (attribute table) and linked to a specific visual attention point, making it easier to combine a quantitative assessment of the visual experience with specific factors such as, for example, user age and lighting conditions.

Despite the important results achieved by visibility studies in the emerging literature mentioned above, a human-centred form of analysis has never been fully accomplished. This workflow introduces a more dynamic approach to overcoming the significant limitations posed by most traditional GIS-based studies.

Conclusions

This research has shown that it is possible to collect, quantify and analyse data concerning the visual experience within a virtually reconstructed Pompeian house. Findings from this newly developed work pipeline may greatly benefit future studies on views and viewing in the Roman house, and the socio-political conditioning of senses as a mode of display for status and power in the ancient domestic realm. Furthermore, it may, optimistically, reach beyond the boundary of the specific area of interest to involve all archaeological contexts in which visual properties and people intertwine, or in which visual properties are manipulated to prompt behaviour.

Acknowledgements

This work is a joint collaboration between the Laboratoriet för Digital Arkeologi (DARKlab, https://www.darklab.lu.se) and the Humanities laboratory (https://www.humlab.lu.se), Lund University, Sweden. The authors sincerely thank the anonymous reviewers. Special gratitude goes to Henrik Gerding for his constructive comments.

Funding statement

This work was supported by Fil dr Uno Otterstedts fond and Bokelunds resestipendiefond.

References

Bek, L. 1980. Towards paradise on Earth: modern space conception in architecture: a creation of Renaissance humanism (Analecta romana Instituti Danici, supplement 9). Odense: Odense University Press.Google Scholar
Campanaro, D.M. 2021. Inference to the Best Explanation (IBE) and archaeology: old tool, new model. European Journal of Archaeology 24: 412–32. https://doi.org/10.1017/eaa.2021.6CrossRefGoogle Scholar
Clarke, J.R. 1991. The houses of Roman Italy, 100 BC–AD 250: ritual, space, and decoration. Berkeley (CA): University of California Press.Google Scholar
Drerup, H. 1959. Bildraum und Realraum im römischen Architektur (Mitteilungen des deutschen archäologischen Instituts: römische Abteilung 66). Heidelberg: F.H. Kerle.Google Scholar
Hales, S. 2003. The Roman house and social identity. Cambridge: Cambridge University Press.Google Scholar
Landeschi, G. et al. 2016. 3D-GIS as a platform for visual analysis: investigating a Pompeian house. Journal of Archaeological Science 65: 103–13. https://doi.org/10.1016/j.jas.2015.11.002CrossRefGoogle Scholar
Llobera, M. 1996. Exploring the topography of mind: GIS, social space and archaeology. Antiquity 70: 612–22. https://doi.org/10.1017/S0003598X00083745CrossRefGoogle Scholar
Opitz, R. 2017. An experiment in using visual attention metrics to think about experience and design choices in past places. Journal of Archaeological Method and Theory 24: 1203–26. https://doi.org/10.1007/s10816-016-9310-2CrossRefGoogle Scholar
Paliou, E. 2013. Reconsidering the concept of visualscapes: recent advances in three-dimensional visibility analysis, in Bevan, A. & Lake, M. (ed.) Computational approaches to archaeological spaces: 243–64. Walnut Creek (CA): Left Coast.Google Scholar
Richards-Rissetto, H. 2017. What can GIS+3D mean for landscape archaeology? Journal of Archaeological Science 84: 10–21. https://doi.org/10.1016/j.jas.2017.05.005CrossRefGoogle Scholar
Trentin, S. 2019. Reality, artifice, and changing landscapes in the house of Marcus Lucretius in Pompeii. Greece and Rome 66: 71–92. https://doi.org/10.1017/S0017383518000323CrossRefGoogle Scholar
Wheatley, D. 1995. Cumulative viewshed analysis: a GIS-based method for investigating intervisibility, and its archaeological application, in Lock, G. & Stancic, Z. (ed.) Archaeology and GIS: a European perspective. London: Taylor & Francis.Google Scholar
Figure 0

Figure 1. The 3D model of the House of the Greek Epigrams, imported in Unity (figure by the authors).

Figure 1

Figure 2. Workflow for the integration of eye-tracking data into a GIS environment (figure by the authors).

Figure 2

Figure 3. Example of light metrics. Level of illumination (LUX) in the reconstructed house of the Greek Epigrams, calculated for four different hours of the winter solstice (unpublished data) (figure by the authors).

Figure 3

Figure 4. JSON datasets exported from Cognitive3D, containing results of a single user's visual experience (a). Eye-tracking-derived datasets are edited with a text editor in order to adjust x,y,z column placements, according to the ArcGIS PRO combination scheme (b).

Figure 4

Figure 5. Exported data are visualised in ArcGIS PRO as point shapefiles. The path of the user exploring the virtual environment of the reconstructed Pompeian house is displayed in the right spatial relation with the original building (figure by the authors).

Figure 5

Figure 6. Gaze and fixation is shown with respect to the surrounding space. Max fixation radius is displayed with graduated symbols based on the size of the radius (figure by the authors).