Author:

Daniel Marius Haselbeck
Supervisor:Prof. Gudrun Klinker
Advisor:David Plecher (@ne23mux)
Submission Date:05/16/2024


Abstract

This project involves Ancient Roman sculpture portraiture, Photogrammetry, 3D object manipulation, the developing and testing of some simple algorithms for comparing 3D objects, and the creation of an android application using Augmented Reality. We compare two ancient statues to each other, one of them showing signs that it was reworked from a different looking original and the other one just being an ordinary statue. The hypothesis we investigate is whether the second statue can serve as an accurate stand-in for the first statue’s original look. To do this, we analyse the statues directly in person and perform 3D scans using images from a smartphone camera. Then, after some postprocessing steps, we also analyse these statues digitally and look for a way to superimpose the ordinary statue on top of the reworked one. We also test a hypothesis about the similarity of areas on the first statue that were possibly left untouched during its rework and the corresponding areas on the second statue using a simple algorithmic approach. Finally, we create an AR phone application to showcase the results from this analysis and place them into the historical context.

Conclusion

We managed to create scans of both statues that are of sufficient quality both for AR-tracking and for usage in our application. We could not produce a definitive connection between the two statues beyond what was already known beforehand. However, our superimposition served well for hightlighting the irregularities of the reworked statue by contrasting them with the normal details of the unedited one. For future digital/algorithmic analyses, we propose starting with original statues of the same typus and focusing on the areas that would stay untouched during a rework. Such approaches should then extrapolate onto reworked statues as well.


[ Thesis ]

 

[ Slides Final ] 

[3D Scans]


Caligula_obj.zipClaudius_obj.zip