Methods for the Virtual Unfolding of Thin-Layer Structures
Abstract:
Thin-layer structures are present in an abondance of objects. Their analysis is often
hampered by the fact that they are usually occluded by other structures or the thin-layer structure
itself. The latter is particularly true, if the thin-layer structure is not flat but strongly curved or even
rolled or folded. The occlusion-free visualization and analysis of such structures require their
extraction and, ideally, their flattening into the 2-dimensional (2D) plane. This process can be
performed in three main steps: (1) generation of a geometric represention of the thin-layer
structure with a 2D manifold surface; (2) flattening of this 2D manifold surface into the 2D plane
with the aim to introduce as little distortions as possible; (3) volumetric transformation of the
image data or direct visualization of the flattened thin-layer volume.

Over the past years, we have substantially contributed to the field of virtual unfolding of thin-layer
structures with innovative approaches. In this talk, I will present our approaches that address all
three processing steps. The feasibility of the approaches will be supported by presenting results
for several thin-layer structures such as rolled and folded ancient documents like papyrus as well
as silver and lead sheets, but also batteries and even biological structures. The talk will be
concluded by an analysis of some remaining challenges.
Lecturer:
Dr. rer. nat. Daniel Baum, Zuse Institute Berlin
Dates:
10.01.2025
Additional Information:
> Lecture website <