A novel network architecture for strip steel surface defect detection was proposed, which designed an enhanced deformable convolution module based on co‑attention mechanism and integrated it into the backbone network as a plugin. By leveraging the co‑attention mechanism, the shape of the convolution kernel was adaptively adjusted, which effectively captured irregular defects on the strip steel surface, and significantly improved the feature extraction capability of the backbone network. Experimental results on the NEU‑DET dataset demonstrated that the proposed method achieved an average precision (mAP) of 81.6%.
XUKe, YANGZhaolin, ZHOUPeng. Technology of online surface inspection for hot rolled steel strips and its industrial application[J]. Journal of Mechanical Engineering, 2009, 45(4): 111‑114, 124.
MAYanting, ZHAOHongdong, YANChao, et al. Strip steel surface defect detection method by improved YOLOv5 network[J]. Journal of Electronic Measurement and Instrumentation, 2022, 36(8): 150‑157.
[5]
WANGShuai, XIAXiaojun, YELanqing, et al. Automatic detection and classification of steel surface defect using deep convolutional neural networks[J]. Metals, 2021, 11(3): 388.
[6]
RENShaoqing, HEKaiming, GIRSHICKR, et al. Faster R‑CNN: Towards real-time object detection with region proposal networks[J]. IEEE Transactions on Pattern Analysis and Machine Intelligence, 2017, 39(6): 1137‑1149.
[7]
HEKaiming, ZHANGXiangyu, RENShaoqing, et al. Deep residual learning for image recognition[C]//2016 IEEE Conference on Computer Vision and Pattern Recognition (CVPR). Las Vegas, 2016: 770‑778.
[8]
LIUXiaoyue, GAOJie. Surface defect detection method of hot rolling strip based on improved SSD model[C]//Database Systems for Advanced Applications. DASFAA 2021 International Workshops, Taipei, 2021: 209‑222.
[9]
LIUWei, ANGUELOVD, ERHAND, et al. SSD: Single shot MultiBox detector[C]//Computer Vision‑ECCV 2016, Amsterdam, 2016: 21‑7.
[10]
DUYongzhao, CHENHaixin, FUYuqing, et al. AFF‑Net: A strip steel surface defect detection network via adaptive focusing features[J]. IEEE Transactions on Instrumentation and Measurement, 2024, 73: 1‑14.
[11]
DAIJifeng, QIHaozhi, XIONGYuwen, et al. Deformable convolutional networks[C]//2017 IEEE International Conference on Computer Vision (ICCV), Venice, 2017: 764‑773.
LIUFengchun, ZHANGJing, XUETao, et al. YOLO‑VDCW:A new light weight surface defect detectional gorithm for stripsteel[J]. China Metallurgy, 2024,34(6):125‑135.
[14]
LEES H, LINS R, CHENS F. Identification of tea foliar diseases and pest damage under practical field conditions using a convolutional neural network[J]. Plant Pathology, 2020, 69(9): 1731‑1739.
[15]
SONGKechen, YANYunhui. A noise robust method based on completed local binary patterns for hot rolled steel strip surface defects[J]. Applied Surface Science, 2013, 285(Part B): 858‑864.
QIUTianheng, WANGLing, WANGPeng, et al. Research on object detection algorithm based on improved YOLOv5[J].Computer Engineering and Applications, 2022,58(13):63‑73.
JIANGJie, ZHANGLimin, LIUKai,et al.Research on infrared ship target detection and tracking method based on improved PP-YOLOE and ByteTrack algorithms[J].Journal of Ordnance Equipment Engineering, 2024, 45(11): 291‑297.
ZHANGTilang, ZHOUMengying, LUYuan, et al. Research on defect detection algorithm of urban drainage network based on improved RTMDet-ins[J]. Water & Wastewater Engineering, 2025, 51(3): 147‑154.
[24]
CARIONN, MASSAF, SYNNAEVEG, et al. End to end object detection with transformers[C]//Computer Vision-ECCV 2020: 16th European Conference, Glasgow, 2020: 213‑229.
[25]
ZHAOYian, WenyuLYU, XUShangliang, et al. Detrs beat YOLOs on real‑time object detection[C]//2024 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), Seattle, 2024: 16965‑16974.