A Novel Algorithm for a Steel Rolling Smart Production using Machine Learning Based Techniques

Abstract

newlineSteel is crucial in various industries due to its durability and versatility. Over the last twodecades, machine vision has become instrumental in enhancing steel product quality throughadvanced surface flaw detection. This technology uses cameras and software algorithms toinspect steel surfaces in real-time, identifying imperfections such as cracks, scratches, and dentsthat could compromise the integrity of the product. The adoption of machine vision for surfaceflaw detection offers a non-contact, highly accurate, and efficient method to ensure steel productsmeet stringent quality standards. This advancement not only improves product reliability but alsosignificantly reduces the cost and time associated with manual inspections, contributing to moresustainable manufacturing practices and higher consumer trust in steel-based structures andproducts. Important problems such as small samples and real-time detection of steel surfacedefects are discussed. Finally, there is the challenge of steel surface flaw detection andthepotential for growth trends.