A deep learning based methodology for artefact identification and suppression with application to ultrasonic images

This paper proposes a deep learning framework for artefact identification and suppression in the context of non-destructive evaluation. The model, based on concept autoencoders, is developed enhancing ultrasound inspection defect through images obtained from full matrix capture data total focusing method. An experimental case study used to prove effectiveness method while exploring its practical limitations. A comparison with state-of-the-art methodology image analysis addressed artefacts. In general, proposed efficiently provides accurate artefacts complex scenarios, even when located below footprint ultrasonic probe, also yields physical parameters needed imaging as by-product. • proposed. grey-box model imaging. trained using defect-free tested data. Ultrasonic interpretation facilitated by removing raw