University of Southampton OCS (beta), CAA 2012

Font Size: 
Exploring the perception of epigraphy: Analysis of visualisations through eye tracking
Adam Frost

Last modified: 2011-12-17


This paper will explore the implementation of 3D reconstruction and eye tracking technology as a quantitative approach to analysing the perception of inscriptions as material culture. Epigraphy is often confined to textual studies as sources of written content, and rarely considered within the built environment, particularly as something that constitutes an element of the visual landscape which is subject to physical factors including visibility and legibility. A novel methodology for the study of the perception of Roman epigraphy is presented, focussing on the use of eye tracking analysis integrated with statistical hypothesis testing to identify how factors such as lighting, relative location and whether the inscription was pigmented affect their visual impact, and its affect on visual attention.

Within the paper, eye tracking is the analysis stage of a workflow that draws on pre-existing pipelines for the production of 3D archaeological reconstructions, including data acquisition, modelling and physically accurate rendering. Eye tracking works by sequentially recording the fixation points of an observer, producing versatile datasets showing areas that were focussed on. It was chosen as a unique method to study human perception within a scientific and quantitative approach, which is described with regard to the important caveats of relating this data to concepts of perception. This paper uses the Roman forum at Wroxeter in Shropshire as a case study in which the entire process is presented, from laser scanning the monumental inscription, modelling and rendering the building, through to the formulation of an experimental design in which 10 participants took part in the eye tracking analysis, concluding with the results of the subsequent statistical hypothesis tests.


Wroxeter; Roman; Forum; Inscription; Eye tracking; Laser scanning; Statistical hypothesis testing;