Automated crack detection on metallic materials with flying-spot thermography using deep learning and progressive training

Helvig, Kevin; Trouvé-Peloux, Pauline; Gaverina, Ludovic; Abeloos, Baptiste; Roche, Jean-Michel

doi:10.1080/17686733.2023.2266176

Cited by 4 publications

(2 citation statements)

References 17 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…To overcome this issue, a curriculum learning approach has been proposed, leveraging on the successive training of a detection network using simulated, synthetic, and real data. 11 Another approach consists in using database for neural network pretraining: an FST dataset for classification acquired on a set of fatigue test metallic specimens with various experimental conditions have been build and released to the community. 12 This database, named FLYD, is built upon reconstructed thermal maps from the recordings on the scans in IR spectrum.…”

Section: Review Of the Literaturementioning

confidence: 99%

“…In the longer run and concerning all the architectures evaluated, the FLYD dataset could be incorporated in a more complex pre-training pipeline, associated with simulated and synthetic data to increase robustness to new defects and noisy data, such as in the curriculum learning approach developped in one of our previous works, originally for classification task. 11…”

Section: Augmentations and Trainingmentioning

confidence: 99%

See 1 more Smart Citation

Database for transfer learning in crack detection and localization on metallic materials using flying spot thermography and deep learning

Helvig,

Trouvé-Peloux,

Gavérina

et al. 2024

J. Electron. Imag.

View full text Add to dashboard Cite

Flying spot" laser infrared thermography (FST) is a non destructive testing technique able to detect defects by scanning surfaces with a laser heat source. Defects, such as cracks on metallic parts, are revealed by the disturbance of heat propagation measured by an infrared camera. Deep learning approaches are now very efficient to automatically analyze and use contextual information from data and can be used for crack detection. However, in the literature only few works deal with the use of deep learning for the crack detection in FST. Indeed obtaining a large amount of data from FST examinations can be expensive and time-consuming. This work presents an open-access database for "flying spot" laser thermography, annotated for crack-type defect localization. It is followed by a benchmark of several state-of-the-art machine learning architectures. The benefits of this database for transfer to the detection and localization of cracks in coated metallic samples are highlighted, with very reduced amounts of data. Finally, this work gives a preliminary exploration on the use of individual thermal images, yielding to detection and localization performance comparable to the previously studied reconstructed thermal maps, which consists in the normalized mean of all the thermal images on a region of interest, after registration.

show abstract

Section: Review Of the Literaturementioning

confidence: 99%

Section: Augmentations and Trainingmentioning

confidence: 99%