Automated defect detection and evaluation in X-ray CT images

X-ray computed tomography gains increasing popularity for industrial quality inspection tasks. It is a challenge however to use it in a fully automated assembly line, which requires robust algorithms for volumetric image analysis on noisy data. This work provides methods for the automatic detection of defects and their geometric description. The approach will be based on spatial statistical analysis of the voxel structure to determine the presence of a defect. Once detected, defect voxels will be clustered to determine shape and size of the associated material faults. The effectiveness of the method will be shown on pores and cracks in steel parts.

Level thesis:degree_level

thesis.doctoral

Grantor dc:publisher

Universität Heidelberg

Year

2002

Author dc:creator

• Eisele, Heiko

Contributors dc:contributor

• Hamprecht, Fred

Repository record source_url

http://www.ub.uni-heidelberg.de/archiv/3106

OAI identifier oai:identifier

oai:archiv.ub.uni-heidelberg.de:3106