DLF-YOLOF:an improved YOLOF-based surface defect detection for steel plate
作者机构：School of Electronic and Information EngineeringUniversity of Science and Technology LiaoningAnshan114051LiaoningChina
出 版 物：《Journal of Iron and Steel Research International》 (J. Iron Steel Res. Int.)
年 卷 期：2024年第31卷第2期
页 面：442-451页
核心收录：
学科分类：08[工学] 080502[工学-材料学] 0805[工学-材料科学与工程（可授工学、理学学位）]
基 金：supported by the Natural Science Foundation of Liaoning Province(No.2022-MS-353) Basic Scientific Research Project of Education Department of Liaoning Province(Nos.2020LNZD06 and LJKMZ20220640)
主 题：Steel surface defects detection YOLOF Anchor-free detector Small object detection Real-time detection
摘 要：Surface defects can affect the quality of steel *** methods based on computer vision are currently applied to surface defect detection of steel ***,their real-time performance and object detection of small defect are still *** improved object detection network based on You Only Look One-level Feature(YOLOF)is proposed to show excellent performance in surface defect detection of steel plate,called ***,the anchor-free detector is used to reduce the network ***,deformable convolution network and local spatial attention module are introduced into the feature extraction network to increase the contextual information in the feature ***,the soft non-maximum suppression is used to improve detection accuracy ***,data augmentation is performed for small defect objects during training to improve detection *** show the average precision and average precision for small objects are 42.7%and 33.5%at a detection speed of 62 frames per second on a single GPU,*** shows that DLF-YOLOF has excellent performance to meet the needs of industrial real-time detection.
维普期刊数据库