Deep Learning-Based Defect Detection System in Steel Sheet Surfaces

Steel is one of the most important building materials of modern times and the production process of flat sheet steel is complicated. Before shipping or delivering steel, sheets need to undergo a careful inspection procedure to avoid defects and thus localizing and classifying surface defects on a steel sheet is crucial. In this study, we advance the steel defect inspection methods by designing machine learning models that aim to detect multi-level defects from sample steel sheet images and classify them according to their corresponding classes. We explore two (2) deep learning methods including U-NET and Deep Residual U-NET to solve the steel defect detection problem with a Dice coefficient accuracy of 0.543 and .731 correspondingly.