Scanning Obsidian: The CAA Conference in Oslo, Norway

Scanning Obsidian: The CAA Conference in Oslo, Norway

I just got back from a whirlwind week in Oslo for the annual Computer Applications and Quantitative Methods in Archaeology conference. I was lucky enough to attend the meetings last year in Sienna, and was super excited to see what was new in the world of applying and teaching digital archaeology this year. The trip included both great discussions with colleagues, and some fun excursions (see the embarrassing selfie from the Viking Ship Museum below).

ship

My own presentation, co-authored with Kele Missal and Leszek Pawlowicz, investigate the best ways to image artifacts made of obsidian (volcanic glass). Obsidian is hard to photograph and 3D scan because it is shiny, and sometimes transparent. The bulk of our study focused on different coatings one can apply to obsidian to make it easier to scan. We also compared how well these different coatings work with different scanning methods (photogrammetry using Agisoft PhotoScan and structured-light scanning using a David SLS-2).

coatings
The coatings we tested in our study.

Our best results were achieved by using brushed talc baby powder on the objects, then modeling them using photogrammetry. Talc powder has other benefits including being cheap and easily available, and being easy to wash off.

By using a two-stage process, taking photos of both a clean object and a coated one, we were also able to produce models with photo-realistic textures.

The second part of our presentation involved the application of reflectance transformation imaging, or RTI. This can be used both for creating amazing two and three dimensional visualizations of objects’ surfaces.

Two products created using normals data generated from RTI. On the left, a gray scale visualization of a projectile point's surface. On the right, a 3D mesh.
Two products created using the normals data generated using RTI. On the left, a grayscale visualization of a projectile point’s surface. On the right, a 3D mesh.

For more information on building a RTI dome, and using it to create different kinds of visualizations, definitely check out Leszek’s site rtimage.us. I hope to add a more detailed post about my own RTI rig up here soon.

My collapsible laser-cut RTI arm.
My collapsible laser-cut RTI arm.

If you’re interested in learning more, you can download our presentation in the form of a pdf (including slides and text) here. We’ve also got a selection of our scans up on Sketchfab. The session our paper was part of (Computer tools for depicting shape and detail in 3D archaeological models) was also recorded, and should be available online shortly, which I’ll of course share here. Later this year, we aim to refine our study a bit more and publish the full results.

To those of you currently at the SAA meetings in Orlando, enjoy the weather! It’s of course snowing in Minneapolis.

Leave a Reply

Your email address will not be published. Required fields are marked *