The steel strip is an essential raw material in the machinery industry. Besides, the surface defects of the steel strip directly determine its performance. To achieve rapid and effective detection of the defects, a CP-YOLOv3-dense (classification priority YOLOv3 DenseNet) neural network was proposed in the present study. The model used YOLOv3 as basic network, implemented priority classification on the images, and then replaced the two residual network modules with two dense network modules. Therefore, the network can receive the multi-layer convolution features output by the dense connection block before making predictions, consequently enhancing the reuse and fusion of features. Finally, the six kinds of surface defects were detected by the improved network. According to the results, the detection precision of the CP-YOLOv3-dense network is 85.7%, the recall rate is 82.3%, the mean average precision is 82.73%, and the detection time of each image is 9.68 ms.
Get full access to this article
View all access options for this article.
References
1.
HeYSongKDongHet al.Semi-supervised defect classification of steel surface based on multi-training and generative adversarial network. Opt Lasers Eng.2019;122:294–302. DOI10.1016/j.optlaseng.2019.06.020.
2.
LiuYXuKWangD.Online surface defect identification of cold rolled strips based on local binary pattern and extreme learning machine. Metals.2018;8:197. DOI:10.3390/met8030197.
3.
NeogiNMohantaDKDuttaPK.Review of vision-based steel surface inspection systems. J Image Video Process. 2014;2014:50. DOI:10.1186/1687-5281-2014-50.
4.
WeimerDScholz-ReiterBShpitalniM.Design of deep convolutional neural network architectures for automated feature extraction in industrial inspection. CIRP Ann Manuf Technols.2016;65:417–420. DOI:10.1016/j.cirp.2016.04.072.
5.
ParkJKKwonBKParkJHet al.Machine learning-based imaging system for surface defect inspection. Int J Precis Eng Manuf -Green Technol.2016;3:303–310. DOI:10.1007/s40684-016-0039-x.
6.
RavikumarSRamachandranKISugumaranV.Machine learning approach for automated visual inspection of machine components. Expert. Syst. Appl.2011;38:3260–3266. DOI:10.1016/j.eswa.2010.09.012.
7.
LuoQFangXSunYet al.Surface defect classification for Hot-rolled steel strips by selectively dominant local binary patterns. IEEE Access2019;7:23488–23499. DOI:10.1109/ACCESS.2019.2898215.
LiJSuZGengJet al.Real-time detection of steel strip surface defects based on improved yolo detection network. IFAC. 2018;51:76–81. DOI:10.1016/j.ifacol.2018.09.412.
10.
RenQGengJLiJ.Slighter Faster R-CNN for real-time detection of steel strip surface defects. 2018 Chinese Automation Congress (CAC);2018; Xi'an, China. p. 2173–2178. DOI:10.1109/CAC.2018.8623407.
11.
LinCYChenCHYangCYet al.Cascading convolutional neural network for steel surface defect detection. Int Conf Appl Hum Factors Ergon. 2019;965:202–212. DOI:10.1007/978-3-030-20454-9_20.
12.
HeDXuKZhouP.Defect detection of hot rolled steels with a new object detection framework called classification priority network. Comput Ind Eng.2019;128:290–297. DOI.10.1016/j.cie.2018.12.043.
13.
TaoXZhangDMaWet al.Automatic metallic surface defect detection and recognition with convolutional neural networks. Appl Sci.2018;8:1575. DOI:10.3390/app8091575.
14.
DongHSongKHeYet al.PGA-Net: pyramid feature fusion and global context attention network for automated surface defect detection. IEEE Transactions on Industrial Informatics. 2019. DOI: 10.1109/TII.2019.2958826.
Saludes-RodilSBaeyensERodríguez-JuanCP.Unsupervised classification of surface defects in wire rod production obtained by eddy current sensors. Sensors. 2015;15:10100–10117. DOI:10.3390/s150510100.
17.
ChenDLiuZ.Generalized Haar-like features for fast face detection. International Conference on Machine learning Cyberne, Hong Kong; 2007. p. 2131–2135. DOI:10.1109/ICMLC.2007.4370496.
18.
HanPLiaoJ.Face detection based on adaboost. International Conference Apperceiving Computing and Intelligence Analysis; Chengdu; 2009. p. 337–340. DOI:10.1109/ICACIA.2009.5361085.
19.
WangZJiaYHuangHet al.Pedestrian detection using Boosted HOG features. Proceedings of the IEEE Conference Intelligence and Transport Systems; Beijing; 2008. p. 1155–1160. DOI:10.1109/ITSC.2008.4732553.
20.
GanGChengJ.Pedestrian detection based on HOG-LBP feature. International Conference and Computing Intelligence Security, Hainan; 2011. p. 1184–1187. DOI:10.1109/CIS.2011.262.
21.
BauerSKöhlerSDollKet al.FPGA-GPU architecture for kernel SVM pedestrian detection. IEEE Computer Society Conference on Computer Vision Pattern Recognition Workshops (CVPRW), 2010. p. 61-68. DOI:10.1109/CVPRW.2010.5543772.
22.
GuoYLiuYOerlemansAet al.Deep learning for visual understanding: a review. Neurocomputing. 2016;187:27–48. doi: 10.1016/j.neucom.2015.09.116
23.
ZhaoZZhengPXuSet al.Object detection with deep learning: a review. IEEE Trans Neural Netw Learn Syst. 2019;30:3212–3232. DOI:10.1109/TNNLS.2018.2876865.
24.
DengLYangMLiTet al.RFBNet: deep multimodal networks with residual fusion blocks for RGB-D semantic segmentation. 2019, arXiv:1907.00135.
25.
DuanKBaiSXieLet al.Centernet: Keypoint Triplets for object detection. 2019 IEEE/CVF International Conference on Computer vision (ICCV); Seoul, Korea (South); 2019. p. 6568–6577.
26.
LawHDengJ.Cornernet: detecting objects as paired keypoints. The European Conference on Computer Vision (ECCV); 2018. p. 734–750.
27.
RenFHuangZ.Facial expression recognition based on AAM–SIFT and adaptive regional weighting. IEEJ Trans Electr Electron Eng. 2015;10:713–722. doi: 10.1002/tee.22151
28.
KasarMMBhattacharyyaDKimTH.Face recognition using neural network: a review. Int J Secur Appl. 2016;10:81–100.
29.
RenFLiuWWuG.Feature reuse residual networks for insect pest recognition. IEEE Access. 2019;7:122758–122768. DOI:10.1109/ACCESS.2019.2938194.
30.
LiuWWuGRenF.Stochastic channel reuse residual networks for plant disease severity detection. IEEE International Conference on Cloud Computing Intelligence Systems (CCIS); Singapore; 2019. p. 57-61. DOI:10.1109/CCIS48116.2019.9073714.
31.
ZhangJZhangS.Error analysis and feature detection of electrochemical machining micro-hole. J Mech Electr Eng. 2019;36:32–35. DOI:10.3969/j.issn.1001-4551.2019.01.007 doi: 10.3901/JME.2000.01.032
32.
RenFZhouY.CGMVQA: a new classification and generative model for medical visual question answering. IEEE Access. 2020;8:50626–50636. DOI:10.1109/ACCESS.2020.2980024.
33.
LatifSUsmanMRanaRet al.Phonocardiographic sensing using deep learning for abnormal heartbeat detection. IEEE Sens J. 2018;18:9393–9400. DOI:10.1109/JSEN.2018.2870759.
34.
ZhangJRenFNiHet al.Research on information recognition of VAT invoice based on computer vision. IEEE International Conference on Cloud Computing Intelligence System (CCIS); Singapore; 2019. p. 126-130. DOI:10.1109/CCIS48116.2019.9073749.
35.
YuanZZhangZSuHet al.Vision-based defect detection for mobile phone cover glass using deep neural networks. Int J Precis Eng Manuf. 2018;19:801–810. DOI:10.1007/s12541-018-0096-x.
36.
LvYMaLJiangH.A mobile phone screen cover glass defect detection model based on small samples learning. IEEE International Conference on Signal Image Process. (ICSIP); Wuxi, China; 2019. p. 1055–1059. DOI:10.1109/SIPROCESS.2019.8868737.
37.
JeyarajPRSamuel NadarER.Computer vision for automatic detection and classification of fabric defect employing deep learning algorithm. Int J Cloth Sci Technol. 2019;31:510–521. DOI:10.1108/IJCST-11-2018-0135.
38.
LiYZhangDLeeD.Automatic fabric defect detection with a wide-and-compact network. Neurocomputing. 2018;329:329–338. DOI:10.1016/j.neucom.2018.10.070.
39.
LinWLinCChenGet al.Steel surface defects detection based on deep learning. International Conference on Applied Human factors Ergon. (AHFE); Jul. 2018; Orlando, USA, 141–149. DOI:10.1007/978-3-319-94484-5_15
40.
LiKWangXJiL.Application of multi-scale feature fusion and deep learning in detection of steel strip surface defect. International Conference on Artificial Intelligence on Advanced Manufacturing (AIAM); Dublin, Ireland; 2019. p. 656–661. DOI:10.1109/AIAM48774.2019.00136.
41.
WeiRSongYZhangY.Enhanced faster region convolutional neural networks for steel surface defect detection. ISIJ Int. 2020;60:539–545. doi: 10.2355/isijinternational.ISIJINT-2019-335
42.
TangRMaoK.An improved GANs model for steel plate defect detection. International Conference on Communication Network and Artificial Intelligence; Dec. 2019; Guangzhou, China. p. 27–29.
43.
RedmonJDivvalaSGirshickRet al.You only look once: unified, real-time object detection. Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition (CVPR); Las Vegas, NV; 2016. p. 779–788. DOI:10.1109/CVPR.2016.91.
44.
RedmonJFarhadiA.YOLO9000: better, Faster, stronger. Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition. (CVPR); Honolulu, HI; 2017. p. 6517–6525. DOI:10.1109/CVPR.2017.690.
45.
RedmonJFarhadiA.Yolov3: An incremental improvement. arXiv preprint arXiv:1804.02767, 2018.
46.
ShaoSMcAleerSYanRet al.Highly Accurate machine fault diagnosis using deep transfer learning. IEEE Trans Indust Inform. 2019;15:2446–2455. DOI:10.1109/TII.2018.2864759.
47.
HuangGLiuSMaatenLVDet al.Condensenet: an efficient densenet using learned group convolutions. Proceedings of the IEEE Conference on Computer Vision Pattern Recognition (CVPR); Salt Lake City, UT; 2018. p. 2752–2761. DOI:10.1109/CVPR.2018.00291.
48.
TianYYangGWangZet al.Apple detection during different growth stages in orchards using the improved YOLO-V3 model. Comput Electron Agric2019;157:417–426. DOI:10.1016/j.compag.2019.01.012.
49.
RenFZhangQ.An emotion expression extraction method for Chinese microblog sentences. IEEE Access. 2020;8:69244–69255. DOI:10.1109/ACCESS.2020.2985726.
50.
WangJLiQGanJet al.Surface defect detection via entity sparsity pursuit with intrinsic priors. IEEE Trans Indust Inform. 2020;16:141–150. DOI:10.1109/TII.2019.2917522.
51.
RenFXueS.Intention detection based on Siamese neural network with triplet loss. IEEE Access. 2020;8:82242–82254. doi: 10.1109/ACCESS.2020.2991484
52.
SongKYanY.A noise robust method based on completed local binary patterns for hot-rolled steel strip surface defects. Appl Surf Sci.2013;285:858–864. doi: 10.1016/j.apsusc.2013.09.002
53.
AliLZhuCZhouMet al.Early diagnosis of Parkinson’s disease from multiple voice recordings by simultaneous sample and feature selection. Expert Syst Appl.2019;137:22–28. DOI:10.1016/j.eswa.2019.06.052.
54.
AliLZhuCZhangZet al.Automated detection of Parkinson’s disease based on multiple types of sustained phonations using linear discriminant analysis and genetically optimized neural network. IEEE J Transl Eng Health Med. 2019;7:1–10. DOI:10.1109/JTEHM.2019.2940900.
55.
HeYSongKMengQet al.An end-to-end steel surface defect detection approach via fusing multiple hierarchical features. IEEE Trans Instrum Meas2020;69:1493–1504. DOI:10.1109/TIM.2019.2915404.
