Non-destructive inspections are repetitive, require focus, and are prone to human error, motivating automated evaluation methods.In this thesis, two approaches are investigated to automate the evaluation of non-destructive testing data. These approaches are based on deep learning and ultrasound wavefield simulations. The non-destructive testing methods examined include ultrasonic testing, thermography, and X-ray computed tomography. X-ray computed tomography serves as a reference for the other methods.
«
Non-destructive inspections are repetitive, require focus, and are prone to human error, motivating automated evaluation methods.In this thesis, two approaches are investigated to automate the evaluation of non-destructive testing data. These approaches are based on deep learning and ultrasound wavefield simulations. The non-destructive testing methods examined include ultrasonic testing, thermography, and X-ray computed tomography. X-ray computed tomography serves as a reference for the other m...
»