Authors:
Nicolas Lermé
1
and
Petr Dokládal
2
Affiliations:
1
MINES ParisTech, PSL Research University and Université Paris-Sud, France
;
2
MINES ParisTech and PSL Research University, France
Keyword(s):
Segmentation, Registration, Shape Analysis, Clustering, Superconductivity, Cable.
Related
Ontology
Subjects/Areas/Topics:
Applications
;
Computer Vision, Visualization and Computer Graphics
;
Geometry and Modeling
;
Image Understanding
;
Image-Based Modeling
;
Pattern Recognition
;
Shape Representation
;
Software Engineering
Abstract:
Superconducting cables have now become a mature technology for energy transport, high-field magnets (MRI, LHC) and fusion applications (ToreSupra, and eventually ITER and DEMO). The superconductors are extremely brittle and suffer from electrical damages brought by mechanical strain induced by electromagnetic field that they generate. An optimal wiring architecture, obtained by simulation, can limit these damages. However, the simulation is a complex process and needs validation. This validation is performed on real 3D samples by the means of image processing. Within this objective, this paper is, to our best knowledge, the first one to present a method to segment the samples of three types of cables as well as a shape and geometry analysis. Preliminary results are encouraging and intended to be later compared to the simulation results.