Strip Steel Defect Classification Using the Improved GAN and EfficientNet
Published in Applied Artificial Intelligence, 2021
Shengqi Guan, Jiang Chang, Hongyu Shi, Xu Xiao, Zhenhao Li, Xu Wang, Xizhi Wang
Strip steel is one of the important raw materials in automotive, marine, aerospace, and other industries. The quality of strip steel directly affects the final performance of industrial products. In the production process of strip steel, various defects such as holes, scratches, rolling, cracks, and pits will occur due to different raw material sources, different processing techniques, and different rolling equipment (Neogi, Mohanta, and Dutta 2014). These defects in strip steel not only affect the appearance but also reduce the wear resistance, corrosion resistance, fatigue resistance, and other physical properties of industrial products, thus leading to the existence of huge potential safety hazards in industrial products. Therefore, it is of great significance to study the defect detection of strip steel.