Industrial X-ray Image Analysis with Deep Neural Networks Robust to Unexpected Input Data

• Published in: Metals (Basel ) Vol. 12; no. 11; p. 1963
• Main Authors: Lindgren, Erik, Zach, Christopher
• Format: Journal Article
• Language: English
• Published: Basel MDPI AG 01.11.2022
• Subjects: Accuracy; Algorithms; Artificial intelligence; Artificial neural networks; Classification; Datasets; Deep learning; Defects; Dictionaries; Fusion welding; Geometry; Hypothesis testing; Image analysis; Image quality; Inspection; Machine learning; Manufacturing; Manufacturing industry; Neural networks; non-destructive evaluation; Production Technology; Produktionsteknik; Quality control; Quality control equipment; Robustness; Training; weld inspection; X ray inspection; X-ray inspection; deep learning; weld inspection; non-destructive evaluation
• ISSN: 2075-4701; 2075-4701
• DOI: 10.3390/met12111963

X-ray inspection is often an essential part of quality control within quality critical manufacturing industries. Within such industries, X-ray image interpretation is resource intensive and typically conducted by humans. An increased level of automatization would be preferable, and recent advances in artificial intelligence (e.g., deep learning) have been proposed as solutions. However, typically, such solutions are overconfident when subjected to new data far from the training data, so-called out-of-distribution (OOD) data; we claim that safe automatic interpretation of industrial X-ray images, as part of quality control of critical products, requires a robust confidence estimation with respect to OOD data. We explored if such a confidence estimation, an OOD detector, can be achieved by explicit modeling of the training data distribution, and the accepted images. For this, we derived an autoencoder model trained unsupervised on a public dataset with X-ray images of metal fusion welds and synthetic data. We explicitly demonstrate the dangers with a conventional supervised learning-based approach and compare it to the OOD detector. We achieve true positive rates of around 90% at false positive rates of around 0.1% on samples similar to the training data and correctly detect some example OOD data.