Graphics and Archaeology: Interpreting the Past

Snowbird, Utah
20-23 May 2000

 

Presentation:

Capturing and Presenting 3D Information on Early Medieval Sculptured Stones

Presenters:

  • Stuart Jeffrey

Presentation Resources:

Summary:

Stuart presented his work with the Scottish Early Medieval Sculptured Stones Project. The 3D computer component of the work came with the need to update and revise a new academic corpus on these stones (that range in date from the 6th to 9th centuries).

The project itself needed to consider:

  • need for a new academic corpus
  • audience and delivery
  • analysis, management, and conservation
  • three dimensional models as a means
  • reconstruction and rejoining [of fragments]
  • money [looking for economical, practical solutions to address the volume of potential data]

The iconic images are often undecipherable and access (because of their positioning) is often a major issue. (Multisided stones leaning up against a wall in a museum only displays one face if it cannot be moved.)

Some of the surface capture techniques possible are:

  • photogrammetry
  • structured light [to reveal the relief of the stones]
  • LIDAR scanning technology
  • QuickTime VR object movies [quick, simple, cheap]

Photogrammetry is difficult because of the intricate, organic shapes of the objects. (Photogrammetry is more effective with regularly shaped objects).

Landscapes and models are becoming more interesting because the importance of landscape context is growing in the analysis and interpretation of these models. Landscape rendering packages are quite interesting, although there remains a big issue about the capture of GIS data - it needs to be finer than the 10m maps currently available in the public domain.

Horizons, too, are important in capturing enough data to support a visual context. Object movies and panoramas are currently being explored here.

Creating models in context need to consider things like:

  • explicitness of reconstructed segments and depth of original data [what is the original sampling quality]
  • scale and orientation of the artefact
  • process of model generation
  • interpretational opportunities
  • objectification of artefacts
  • academic and public audiences [his audiences are primarily academic]

Conclusions / Issues:

  • Example - Stuart showed a stone cross they had modeled in 3D StudioMax - there was an explicit separation of reconstructed (estimated) elements versus the actual remains - interestingly, the didn't know what the heading/footing looked like, but they believe it to be a cross in their best estimation
  • Example - Stuart showed a QTVR image of a stone fragment - moving the light source in a structured fashion allowed the viewer to see/select views that revealed the most information about the relief of the stone
  • It was noted that there was an overlap in several projects and this study - Lon Addison's remarks about repeating work are important here - it is important that we remain in contact
  • Question - I'm interested in the notion that landscape context was important in what the monuments said about the landscape (safe, dangerous, sacred, etc.) - has there been any effort to capture landscape data and analyze it for patterns?
    • yes, capture of landscapes is of interest - however, estimating where stones (that are now in museums) were originally would be difficult
  • Question - wasn't the Scottish landscape more forested?
    • yes, and this is why computer models would be helpful in estimating vegetation in context

 

[ Previous ] [ Index ] [ Next ]