Deep learning techniques for non-destructive testing and evaluation - CEA - Commissariat à l’énergie atomique et aux énergies alternatives Accéder directement au contenu
Chapitre D'ouvrage Année : 2022

Deep learning techniques for non-destructive testing and evaluation

Résumé

In this chapter, we analyzed some applications of deep learning methods to electromagnetic NDT&E and tried to show how deep neural networks can be adapted to different scenarios involving electromagnetic probing waves ranging from the quasi-static regime to microwave. In particular, CNN have been deeply exploited when the treated signals behave "as images" such as in the case of ECT and MFL inspections where real and imaginary parts of the impedance variation as well as the magnetic flux density are probed. Furthermore, time domain signals as in PECT or GPR measurements have been addressed, too, by employing LSTM-RNN and/or through CNN explicitly adapted for the purpose (e.g., pixel-wise inversion). Our analysis underlined that specifically tailored deep neural architectures have obtained a better prediction performances than pre-trained networks based on state-of-the-art architectures. In fact, the systematic lack of large shared datasets containing labeled measurements of realistic acquisitions makes it difficult to properly benchmark and improve such backbone architectures. Moreover, the difficulties in collecting labeled measurements for defect parameters (e.g., the defect geometry) downsize the practical applications of deep learning models mostly to classification problems. The survey performed in this chapter has also highlighted that the application of deep learning in NDT&E is also going toward the acceleration of numerical forward solvers for NDT&E modeling and simulations in a fully model-driven approach. It is believed that the ability of DL methods to handle problems having large cardinality (e.g., NDT&E parameters such as large number of defect classes, and defect geometry description) will boost the research and its application to time consuming statistical studies (see, e.g., [160,161]). Moreover, our analysis showed that the use of numerical solvers proves useful in designing the most suitable DL schemas as well as in improving the prediction accuracy when a low amount of measurements is available. Finally, a large amount of works in the literature showed that exploitation of deep learning algorithms directly on embedded systems (e.g., FPGA hardware) is already possible without an appreciable degradation in prediction performance.
Fichier principal
Vignette du fichier
Chapter_04_final.pdf (3.8 Mo) Télécharger le fichier
Origine Fichiers produits par l'(les) auteur(s)

Dates et versions

cea-04316587 , version 1 (30-11-2023)

Identifiants

Citer

Roberto Miorelli, Anastassios Skarlatos, Caroline Vienne, Christophe Reboud, Pierre Calmon. Deep learning techniques for non-destructive testing and evaluation. Maokun Li; Marco Salucci. Applications of deep learning in electromagnetics: Teaching Maxwell's equations to machines, Scitech Publishing, pp.99-143, 2022, 978-1839535895. ⟨10.1049/SBEW563E⟩. ⟨cea-04316587⟩
166 Consultations
212 Téléchargements

Altmetric

Partager

Gmail Mastodon Facebook X LinkedIn More