Projekte

Europe must proactively exploit markets in which it has strategic advantages in order to develop products that will make it a leading player. One such area is Cultural Heritage (CH). Having studied the needs of CH for a number of years, PRESIOUS aims to investigate innovative ICT solutions to the following key identified challenges: a) the difficulty and inefficiency of the 3D digitisation process, b) the quantification of stone monument degradation, and c) the reconstruction of objects from large numbers of constituent fragments that may be worn, immovable, dispersed or incomplete. Using a common core of geometric processing, analysis and retrieval methods, PRESIOUS will develop predictive geometric augmentation technologies targeted to the above problems:

- On-the-fly auto-completion for 3D digitization that exploits similar template models to allow gradual shape prediction from partially digitized objects. Such a breakthrough system opens a whole new range of possibilities for decreased acquisition times and simplified procedures, leading to an order of magnitude reduction in effort and cost. This technology is a prime candidate for exploitation by our industrial partner.

- Estimation and prediction of monument degradation. The project will investigate models for forward and inverse erosion prediction based on targeted high-accuracy surface scans, allowing one to essentially move the artefact's surface condition "back and forth in time" and visualize and measure the dynamic state of the deteriorating object. This tool is also likely to be applicable in areas outside CH, such as civil engineering. 3D CH object repair. Automated, predictive reconstruction from fragmented CH objects that goes beyond reassembly by proposing the synthesis of missing parts using geometry auto-completion. This will aid the physical and virtual restoration process.

In PRESIOUS, the University of Konstanz partner is focussing on innovative methods for object repair. Specifically, we will consider novel similarity-based approaches for reassembly and completion of partially represented objects from the 3D Cultural Heritage domain. To this end, appropriate partial retrieval methods need to be developed, which can retrieve the most appropriate models from a template shape repository for repair and completion purposes.

• Schreck, Tobias - Projektleiter*in

• FB Informatik und Informationswissenschaft