Surface defect diagnosis for steel sheets is essential to maintain the quality of the product. However, it is challenging to do such a task due to the high manufacturing speed of 100–140 m/min, making it impossible to observe defects visually. However, developing a lightweight and real-time adaptable intelligent approach can both overcome this issue and satisfy the requirements of the concept of Industry 4.0. Based on that motivation, this study aimed to satisfy such a demand by introducing an attention-based surface defect identification approach called MobileNetV2-AGCA and UNET-AGCA for defect localisation. The effectiveness of the proposed techniques was evaluated by considering steel strip surface defect images of the NEU-CLS and NEU-DET surface defect dataset where six different defects are included. The Adaptive Graph Channel Attention (AGCA) Module was implemented in an edited version of MobileNetV2 and UNET architecture to improve segmentation accuracy while maintaining or improving the execution speed. According to the experimental results, MobileNetV2-AGCA successfully identified the surface defects with an average accuracy of 99.72%. Regarding defect localisation, UNET-AGCA reached an F1 score of 0.795. It is concluded that the proposed methods provided promising results to be applied to a real-time problem related to surface defect identification and localisation.
MorganJHaltonMQiaoY, et al.Industry 4.0 smart reconfigurable manufacturing machines. J of Manuf Syst2021; 59: 481–506.
2.
ChauhanSSinghRGehlotA, et al.Digitalization of supply chain management with industry 4.0 enabling technologies: a sustainable perspective. Processes2022; 11: 96.
3.
DimitrovaMAminzadehAMeiabadiMS, et al.A survey on non-destructive smart inspection of wind turbine blades based on industry 4.0 strategy. Appl Mech2022; 3: 1299–1326.
4.
IslamMMMKimJ-M. Automated bearing fault diagnosis scheme using 2D representation of wavelet packet transform and deep convolutional neural network. Comput in Ind2019; 106: 142–153.
5.
KimJKimJ-M. Bearing fault diagnosis using grad-CAM and acoustic emission signals. Appl Sci2020; 10: 2050.
6.
TraylorCInalpolatM. A generalized computational approach to predict high-frequency acoustic pressure response of cavity structures for structural health monitoring of wind turbine blades. Wind Eng2021; 46: 0309524X–02110605
7.
ZhuYLiGTangS, et al.Acoustic signal-based fault detection of hydraulic piston pump using a particle swarm optimization enhancement CNN. Appl Acoustics2022; 192: 108718–108718.
8.
QiaoMYanSTangX, et al.Deep convolutional and LSTM recurrent neural networks for rolling bearing fault diagnosis under strong noises and variable loads. IEEE Access2020; 8: 66257–66269.
9.
GanaMAchourHBelaidK, et al.Non-invasive intelligent monitoring system for fault detection in induction motor based on lead-free-piezoelectric sensor using ANN. Measure Sci Technol2022; 33: 065105–065105.
10.
DasOBagci DasD. Smart machine fault diagnostics based on fault specified discrete wavelet transform. J of the Brazilian Soc of Mech Sci Eng2023; 45: 55.
11.
GuoSZhangBYangT, et al.Multitask convolutional neural network with information fusion for bearing fault diagnosis and localization. IEEE Trans on Ind Electron2020; 67: 8005–8015.
12.
HeFYeQ. A bearing fault diagnosis method based on wavelet packet transform and convolutional neural network optimized by simulated annealing algorithm. Sensors2022; 22: 1410.
13.
LiMWangHWanZ. Surface defect detection of steel strips based on improved YOLOv4. Comput Electr Eng2022; 102: 108208.
14.
ChaY-JChoiWBüyüköztürkO. Deep learning-based crack damage detection using convolutional neural networks. Comput-Aided Civil and Infrastruct Eng2017; 32: 361–378.
15.
KonovalenkoIMaruschakPBrezinováJ, et al.Steel surface defect classification using deep residual neural network. Metals2020; 10: 846.
16.
WangSXiaXYeL, et al.Automatic detection and classification of steel surface defect using deep convolutional neural networks. Metals2021; 11: 388.
17.
LiWZhangHWangG, et al.Deep learning based online metallic surface defect detection method for wire and arc additive manufacturing. Robotics and Comput-Integr Manuf2023; 80: 102470.
18.
ZhouDKangBJinX, et al.DeepViT: towards deeper vision transformer. arXiv:210311886 [cs], https://arxiv.org/abs/2103.11886 (2021, accessed 15 August 2024).
19.
ChenC-FFanQPandaR.CrossViT: cross-attention multi-scale vision transformer for image classification. arXiv:210314899 [cs], https://arxiv.org/abs/2103.14899 (2021, accessed 15 August 2024).
20.
MehtaSRastegariM.MobileViT: light-weight, general-purpose, and mobile-friendly vision transformer. arXiv:211002178 [cs], https://arxiv.org/abs/2110.02178 (2022, accessed 15 August 2024).
HeHYuanGZhouH, et al.Solar radio burst detection based on the MobileViT-SSDLite lightweight model. The Astrophysical J Supplement Series2023; 269: 51–51.
23.
DingYYiZLiM, et al.HI-MViT: a lightweight model for explainable skin disease classification based on modified MobileViT. Digital Health2023; 9. DOI: 10.1177/20552076231207
24.
YaoHLiuTLiG, et al.Screen-shot and demoiréd image identification based on DenseNet and DeepViT. Expert Syst with Appl2024; 240: 122580–122580.
25.
ZhengPwangHSangZ, et al.Smart manufacturing systems for industry 4.0: conceptual framework, scenarios, and future perspectives. Frontiers of Mech Eng2018; 13: 137–150.
26.
GogolinskiyKVSyaskoVA. Prospects and challenges of the fourth industrial revolution for instrument engineering and metrology in the field of non-destructive testing and condition monitoring. Insight - Non-Destruct Test Cond Monitoring2019; 61: 434–440.
27.
JaberAKarganroudiSS, Saleh, et al.On smart geometric non-destructive evaluation: inspection methods, overview, and challenges. Materials2022; 15: 7187–7187.
28.
RussakovskyODengJSuH, et al.Imagenet large scale visual recognition challenge. Int J of Comput Vis2015; 115: 211–252.
29.
ZhangKZuoWChenY, et al.Beyond a Gaussian denoiser: residual learning of deep CNN for image denoising. IEEE Trans on Image Process2017; 26: 3142–3155.
30.
KrizhevskyASutskeverIHintonGE. Imagenet classification with deep convolutional neural networks. Communications of the ACM2012; 60: 84–90.
31.
ZhaoJLiuSTaoH, et al.Slim-YOLOv8: a fast and accurate algorithm for surface defect detection of steel strips. Ironmaking & Steelmaking2024. DOI: 10.1177/03019233241266717
32.
XueBWuZ. Key technologies of steel plate surface defect detection system based on artificial intelligence machine vision. Wirel Commun Mob Comput2021; 2021: 1–12. DOI: 10.1155/2021/5553470
33.
CuiLJiangXXuM, et al.SDDNet: a fast and accurate network for surface defect detection. IEEE Trans on Instrum Measure2021; 70: 1–13.
34.
WenHLiYWangY, et al.Usage of an improved YOLOv5 for steel surface defect detection. Mater Testing/Materialprüfung2024; 66: 726–735.
35.
TaoXZhangDMaW, et al.Automatic metallic surface defect detection and recognition with convolutional neural networks. Appl Sci2018; 8: 1575.
36.
HeYSongKMengQ, et al.An end-to-end steel surface defect detection approach via fusing multiple hierarchical features. IEEE Trans on Instrum Measure2020; 69: 1493–1504.
37.
CuiWSongKFengH, et al.Autocorrelation-aware aggregation network for salient object detection of strip steel surface defects. IEEE Trans on Instrum Measure2023; 72: 5019412.
38.
FengHSongKCuiW, et al.Cross position aggregation network for few-shot strip steel surface defect segmentation. IEEE Trans on Instrum Measure2023; 72: 5007410.
39.
LiuHChenCHuR, et al.CGTD-Net: channel-wise global transformer-based dual-branch network for industrial strip steel surface defect detection. IEEE Sens J2024; 24: 4863–4872.
40.
LiZWeiXHassaballahM, et al.A deep learning model for steel surface defect detection. Complex & Intell Syst2024; 10: 885–897.
41.
ZhouPZhouGLiY, et al.A hybrid data-driven method for wire rope surface defect detection. IEEE Sens J2020; 20: 8297–8306.
42.
LiuKLuoNLiA, et al.A new self-reference image decomposition algorithm for strip steel surface defect detection. IEEE Trans on Instrum Measure2020; 69: 4732–4741.
43.
LvXDuanFJiangJ, et al.Deep metallic surface defect detection: the new benchmark and detection network. Sensors2020; 20: 1562.
44.
HaoRLuBChengY, et al.A steel surface defect inspection approach towards smart industrial monitoring. J of Intell Manuf2020; 32: 1833–1843.
45.
YuJGaoHSunJ, et al.Automatic and efficient metallic surface defect detection based on key pixel point locations. IEEE Sens J2021; 21: 11476–11487.
46.
LiZWuCHanQ, et al.CASI-Net: a novel and effect steel surface defect classification method based on coordinate attention and self-interaction mechanism. Mathematics2022; 10: 963.
47.
DemirKAyMCavasM, et al.Automated steel surface defect detection and classification using a new deep learning-based approach. Neural Comput Appl2022; 35: 8389–8406Epub ahead of print 2022.
48.
GuoCSzemenyeiMYiY, et al.SA-UNet: spatial attention U-net for retinal vessel segmentation. arXiv: 2004.03696 [cs], (2021, accessed 15 August 2024).
49.
XiangXWangZZhangJ, et al.AGCA: an adaptive graph channel attention module for steel surface defect detection. IEEE Trans on Instrum Measure2023; 72: 1–12.
50.
LundbergS-E.Development trends in high productivity hot rolling of high quality high carbon steel wire rod. BHM : Berg- und Hüttenmännische Monatshefte2002;147:333–344.
51.
GuoJLiuPXiaoB, et al.Surface defect detection of civil structures using images: review from data perspective. Autom in Constr2024; 158: 105186.
52.
SandlerMHowardAZhuM, et al.MobileNetV2: inverted residuals and linear bottlenecks. arXiv: 1801.04381, https://arxiv.org/abs/1801.04381 (2018, accessed 15 August 2024).
53.
ZhengHChenXChengH, et al.MD-YOLO: surface defect detector for industrial complex environments. Opt Lasers in Eng2024; 178: 108170–108170.
54.
LiuTHeZLinZ, et al.An adaptive image segmentation network for surface defect detection. IEEE Trans on Neural Netw Learn Syst2024; 35: 1–14.
55.
LuJZhuMMaX, et al.Steel strip surface defect detection method based on improved YOLOv5 s. Biomimetics2024; 9: 28–28.
56.
WangHYangXZhouB, et al.Strip surface defect detection algorithm based on YOLOv5. Materials2023; 16: 2811–2811.
57.
QianXWangXYangS, et al.LFF-YOLO: a YOLO algorithm with lightweight feature fusion network for multi-scale defect detection. IEEE Access2022; 10: 130339–130349.
58.
SongXCaoSZhangJ, et al.Steel surface defect detection algorithm based on YOLOv8. Electronics2024; 13: 988–988.
59.
XingJJiaM. A convolutional neural network-based method for workpiece surface defect detection. Measurement2021; 176: 109185.
60.
GuoXGongKLuC. Low-resolution steel surface defects classification network based on autocorrelation semantic enhancement. Coatings2015; 13: 2015. DOI: 10.3390/coatings13122015
61.
JiangXCuiYCuiY, et al.Optimization algorithm of steel surface defect detection based on YOLOv8n-SDEC. IEEE Access2024; 12: 95106–95117.
62.
LiuTHeZLinZ, et al.An adaptive image segmentation network for surface defect detection. IEEE Trans on Neural Netw Learn Syst2024; 35(6): 8510–8522.
63.
LiuTHeZ. TAS2-Net: tripple-attention semantic segmentation network for small surface defect detection. IEEE Trans on Instrum Measure2022; 71: 5004512.
64.
LiWLiBNiuS, et al.LSA-Net: location and shape attention network for automatic surface segmentation. J of Manuf Process2023; 99: 65–77.
65.
YangHWangYHuJ, et al.Segmentation of track surface defects based on machine learning vision and neural networks. IEEE Sens J2022; 22: 1571–1581.
