Florian Gawrilowicz: 3D surfaces are a fundamental part of many applications, including robotics, photogrammetry and virtual reality.
As scanning hardware continues to improve, larger and more complex 3D models will become commonplace and their increasing proliferation will help fuel new applications. With the emergence of mobile scanning, we will soon be swimming in a sea of geometric data. In the light of this increased model complexity and ubiquity, processing methods are forced to keep pace with the explosion of 3D data.
From the nature of digital acquisition all of the available, and presumably upcoming technologies generate a collection of singular points. There is no direct information captured about how these points relate to each other. For most applications, it is necessary or at least advantageous to have a connected surface to work with.
The scope of the project within this context is to utilize the information about the data in a metrological fashion - meaning to be able to state the level of precision and uncertainty of each datum. In the line of that view the notion of traceability is introduced, where the actually measured data is not replaced. This is where the current state-of-the-art in surface reconstruction falls short because they estimate a smoothed surface from the input data and introduce new points corresponding to that estimation. New methods for surface reconstruction are to be explored in the course of the Ph.D. project.
PhD project: A Traceable 3D Scanning and Reconstruction Pipeline
Supervisors: Jakob Andreas Bærentzen; Anders Bjorholm Dahl