Automated Pitting Corrosion Detection of Metallic Glass Using Deep Learning

Metallic glasses are unique alloys with a random atomic structure, offering exceptional strength, hardness, and corrosion resistance. However, they are prone to corrosion in harsh environments, making the study of corrosion mechanisms vital for broadening their applications. Despite extensive research on metallic glasses, there is a gap in utilising Scanning Electron Microscopy (SEM) images and deep learning for detecting pitting corrosion. This study addresses this gap by employing deep learning algorithms—RetinaNet101, Faster Region-based Convolutional Neural Network (Faster RCNN), Mask Region-based Convolutional Neural Network (Mask RCNN), and You Only Look Once version 5 (YOLOv5)—to identify and locate corrosion pits in SEM images of metallic glass. The models are trained and evaluated for precision, recall, and F1 score. Among these, Faster RCNN achieved the highest F1 score of 95.7<sup xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">%</sup>• These findings contribute to advancing corrosion detection techniques, enhancing the practical applications of metallic glasses in various industries.