CASI-Net: A Novel and Effect Steel Surface Defect Classification Method Based on Coordinate Attention and Self-Interaction Mechanism

The surface defects of a hot-rolled strip will adversely affect the appearance and quality industrial products. Therefore, timely identification is great significance. In order to improve efficiency accuracy defect detection, lightweight network based on coordinate attention self-interaction (CASI-Net), which integrates channel domain, spatial information, module, proposed automatically identify six kinds steel defects. this paper, we use embed location information into attention, enables CASI-Net locate region more accurately, thus contributing better recognition classification. addition, features are converted aggregation from horizontal vertical direction attention. Furthermore, module interactively fuse extracted feature classification accuracy. experimental results show that can achieve accurate with reduced parameters computation.