Autoencoder-based Ultrasonic NDT of Adhesive Bonds
- We present an approach for ultrasonic non-destructive testing of adhesive bonding employing unsupervised machine learning with autoencoders.The models are trained exclusively on the features derived from pulse-echo ultrasonic signals on a specimen with good adhesive bonding and tested on another specimen with artificially added defects.The resulting pseudo-probabilities indicating anomalies are visualized and presented along to the C-scan of the same specimen. As a result, we achieved improved representation of the defects, allowing their automatic and reliable detection.
Author: | Ivan Kraljevski, Frank Duckhorn, Martin Barth, Constanze Tschöpe, Frank Schubert, Matthias WolffORCiD |
---|---|
DOI: | https://doi.org/10.1109/SENSORS47087.2021.9639864 |
ISBN: | 978-1-7281-9501-8 |
Title of the source (English): | IEEE SENSORS 2021, Conference Proceedings, Oct 31- Nov 4, Sydney, Australia |
Publisher: | IEEE |
Document Type: | Conference publication peer-reviewed |
Language: | English |
Year of publication: | 2021 |
Number of pages: | 4 |
Faculty/Chair: | Fakultät 1 MINT - Mathematik, Informatik, Physik, Elektro- und Informationstechnik / FG Kommunikationstechnik |