文献类型：期刊文献

英文题名：SDD-YOLO: A Lightweight, High-Generalization Methodology for Real-Time Detection of Strip Surface Defects

作者：Wu, Yueyang[1];Chen, Ruihan[1,2];Li, Zhi[1];Ye, Minhua[3];Dai, Ming[1]

机构：[1]Guangdong Ocean Univ, Sch Math & Comp, Zhanjiang 524008, Peoples R China;[2]Int Macau Inst Acad Res, Artificial Intelligence Res Inst, Taipa 999078, Macau, Peoples R China;[3]Guangdong Ocean Univ, Coll Ocean Engn & Energy, Zhanjiang 524088, Peoples R China

年份：2024

卷号：14

期号：6

外文期刊名：METALS

收录：SCI-EXPANDED(收录号：WOS:001256629500001)、、Scopus(收录号：2-s2.0-85197229061)、WOS

基金：This research was supported by a special grant from the Guangdong Basic and Applied Basic Research Foundation under Grant No. 2023A1515011326, the program for scientific research start-up funds of Guangdong Ocean University under Grant No. 060302102101, Guangdong Provincial Science and Technology Innovation Strategy under Grant No. pdjh2023b0247,National College Students Innovation and Entrepreneurship Training Program under Grant No.202310566022, and Guangdong Ocean University Undergraduate Innovation Team Project under Grant No. CXTD2023014.

语种：英文

外文关键词：strip defect detection; SDD-YOLO; YOLOv5s; BiFPN; lightweight; NEU-DET

外文摘要：Flat-rolled steel sheets are one of the major products of the metal industry. Strip steel's production quality is crucial for the economic and safety aspects of humanity. Addressing the challenges of identifying the surface defects of strip steel in real production environments and low detection efficiency, this study presents an approach for strip defect detection based on YOLOv5s, termed SDD-YOLO. Initially, this study designs the Convolution-GhostNet Hybrid module (CGH) and Multi-Convolution Feature Fusion block (MCFF), effectively reducing computational complexity and enhancing feature extraction efficiency. Subsequently, CARAFE is employed to replace bilinear interpolation upsampling to improve image feature utilization; finally, the Bidirectional Feature Pyramid Network (BiFPN) is introduced to enhance the model's adaptability to targets of different scales. Experimental results demonstrate that, compared to the baseline YOLOv5s, this method achieves a 6.3% increase in mAP50, reaching 76.1% on the Northeastern University Surface Defect Database for Detection (NEU-DET), with parameters and FLOPs of only 3.4MB and 6.4G, respectively, and FPS reaching 121, effectively identifying six types of defects such as Crazing and Inclusion. Furthermore, under the conditions of strong exposure, insufficient brightness, and the addition of Gaussian noise, the model's mAP50 still exceeds 70%, demonstrating the model's strong robustness. In conclusion, the proposed SDD-YOLO in this study features high accuracy, efficiency, and lightweight characteristics, making it applicable in actual production to enhance strip steel production quality and efficiency.