Application of machine learning methods on the defect detection in shearographic images

Publikation: Beitrag in Buch/Konferenzbericht/Sammelband/GutachtenBeitrag in KonferenzbandBeigetragenBegutachtung

Abstract

Defect detection in primary composite lightweight structures is a major ongoing challenge in terms of reliability, rapidity, and accuracy for a secure operational life-cycle. Shearography is a full-field and material independent non-destructive testing method. Despite its major suitability for large composite components, this method still requires specialists to reliably identify the defect patterns. Modern algorithms in terms of machine learning have gained huge popularity and provide the ability to outperform conventional algorithms, especially in image analysis. Hence, an object detection model based on convolutional neural networks has been implemented and applied to shearographic images of different composite specimens. Concluding, the model performs with a considerable high accuracy considering the medium sized and manually labeled dataset.

Details

OriginalspracheEnglisch
TitelProceedings of the 20th European Conference on Composite Materials
Redakteure/-innenAnastasios P. Vassilopoulos, Véronique Michaud
Herausgeber (Verlag)Ecole Polytechnique Fédérale de Lausanne (EPFL)
Seiten492-501
Seitenumfang10
Band3
ISBN (elektronisch)978-2-9701614-0-0
PublikationsstatusVeröffentlicht - 12 Dez. 2022
Peer-Review-StatusJa

Konferenz

Titel20th European Conference on Composite Materials
UntertitelComposites Meet Sustainability
KurztitelECCM 20
Veranstaltungsnummer20
Dauer26 - 30 Juni 2022
Webseite
BekanntheitsgradInternationale Veranstaltung
OrtSwissTech Convention Center
StadtLausanne
LandSchweiz

Externe IDs

Scopus 85149178017
ORCID /0000-0003-1370-064X/work/142243796
ORCID /0000-0002-6817-1020/work/142246611
ORCID /0000-0003-2653-7546/work/142249396

Schlagworte

Schlagwörter

  • shearography, non-destructive testing, composite, machine learning