A defect detection algorithm based on dual-level cascade and multi-feature scale fusion Faster Region based Convolutional Neural Network (Faster-RCNN) is proposed to overcome the challenges of uneven distribution, small size, and diverse morphology of interface surface defects on the automotive brake pipe. Basing on Faster-RCNN, the proposal of multi-scale feature fusion combines feature information from different scales together to ensure that easily overlooked small targets can be noticed. Region of Interest (RoI) alignment pooling is proposed to reduce the deviation between the original feature regions and the mapped feature regions. To reduce the impact of uneven sample distribution on algorithm performance, a two-stage cascade structure is constructed to enable that different samples use different Intersection over Union (IoU) thresholds. The design of the loss function was optimized to balance the learning ability of the algorithm between easy and difficult samples. A “Defect” dataset is constructed and it validates the effectiveness of the improved algorithm. The improved algorithm achieves Precision (P) of 90.3%, Recall (R) of 84.5%, and mean Average Precision (mAP) of 91.1%. It has been demonstrated that the improved Faster-RCNN defect detection algorithm exhibits high accuracy and robustness. According to the production requirements of the actual industrial production, an integrated online defect detection system has been established, which combines testing for rotation integrity, groove completeness, and interface surface defects. The experimental results indicate that the system achieves an accuracy of 99.4%, a false positive rate of 1.2%, a false negative rate of 0.5%, and a detection time of 0.78 s. The integrated online defect detection system reduces the reliance on manual inspection, enhancing production efficiency and ensuring consistent quality, thereby meeting stringent production standards.
StamouGAngelopoulosSKtenaA, et al. 3D Anisotropic Magnetoresistance sensor for steel health monitoring. Eng Fail Anal2024; 160: 108165.
2.
StamouGAngelopoulosSHristoforouE. Development of a steel health monitoring device based on anisotropic magnetoresistance sensors. Int J Appl Electromagn Mech2024; 74(4): 357–362.
3.
WangCCWeiXMJiangXS. An automated defect detection method for optimizing industrial quality inspection. Eng Appl Artif Intell2024; 127: 107387.
4.
ShenSQRenKBaoYQ, et al. Machine vision–based automatic planning for microdroplet coating trajectory on irregularly shaped regions of metal surfaces. J Manuf Process2023; 106: 117–129.
5.
TempelaereADe KetelaereBHeJ, et al. An introduction to artificial intelligence in machine vision for postharvest detection of disorders in horticultural products. Postharvest Biol Technol2023; 206: 112576.
6.
ZhongZFWangGTanJJ, et al. Surface object detection algorithm of lightweight steel rolling based on machine vision. SN Comput Sci2023; 4: 806.
7.
WangBWangMYangJZ, et al. YOLOv5-CD: strip steel surface defect detection method based on coordinate attention and a decoupled head. Meas Sens2023; 30: 100909.
8.
HassanSDhimishM. Enhancing solar photovoltaic modules quality assurance through convolutional neural network-aided automated defect detection. Renew Energy2023; 219: 119389.
9.
DongSHSunXXieSY, et al. Automatic defect identification technology of digital image of pipeline weld. Nat Gas Ind B2019; 6: 399–403.
10.
GaoJXHengFYuanYP, et al. A novel machine learning method for multiaxial fatigue life prediction: improved adaptive neuro-fuzzy inference system. Int J Fatigue2024; 178: 108007.
JianYYuanG. Research on the defects identify of the weld-line’s image of
the in-service pipeline based on BP neural network. In: 2017 first international conference on electronics instrumentation & information systems (EIIS), Harbin, China, 3–5 June 2017, pp.1–6. New York: IEEE.
13.
JiangHQZhaoYLGaoJM, et al. Weld defect classification based on texture features and principal component analysis. Insight2016; 58: 194–200.
14.
OuladbrahimABelaidiIKhatirS, et al. Experimental crack identification of API X70 steel pipeline using improved Artificial Neural Networks based on Whale Optimization Algorithm. Mech Mater2022; 166: 104200.
15.
LimTRatnamMKhalidM. Automatic classification of weld defects using simulated data and an MLP neural network. Insight2007; 49: 154–159.
16.
LiuHGuoRY. Detection and identification of SAWH pipe weld defects based on X-ray image and CNN. Chin J Sci Instrum2018; 39: 247–256.
17.
ChengSYangHGXuXQ. Improved lightweight X-Ray aluminum alloy weld defect detection algorithm based on YOLOv5. Chin J Lasers2022; 49: 2104005.
18.
LiYGaoW. Research on X-ray welding image defect detection based on convolution neural network. J Phys Conf Ser2019; 1237: 032005.
19.
HouWHWeiYGuoJ, et al. Automatic detection of welding defects using deep neural network. J Phys Conf Ser2018; 933: 012006.
20.
SaeediJDottaMGalliA, et al. Measurement and inspection of electrical discharge machined steel surfaces using deep neural networks. Mach Vis Appl2021; 32: 21.
21.
PizarroRAssemlalH-EJegathambalSKB, et al. Deep learning, data ramping, and uncertainty estimation for detecting artifacts in large, imbalanced databases of MRI images. Med Image Anal2023; 90: 102942.
22.
TangFSGaoQDuZZ. Algorithm of object localization applied on high-voltage power transmission lines based on line stereo matching. Opt Eng2021; 60: 023101.
23.
GodalaSKumarMS. A weight optimized deep learning model for cluster based intrusion detection system. Opt Quantum Electron2023; 55: 1224.
24.
RedmonJDivvalaSGirshickR, et al. You only look once: unified, real-time object detection. In: Proceedings of the IEEE conference on computer vision and pattern recognition, Las Vegas, NV, USA, 27–30 June 2016, pp.779–788. New York: IEEE.
25.
LiuWAnguelovDErhanD, et al. SSD: single shot multibox
detector. In: LeibeBMatasJSebeN,
et al. (eds) Computer vision–ECCV 2016. Cham: Springer, 2016, pp.21–37.
26.
LiGJHuangXJAiJY, et al. Lemon-YOLO: an efficient object detection method for lemons in the natural environment. IET Image Process2021; 15: 1998–2009.
27.
LiuBLLuoHWangHT, et al. YOLOv3_ReSAM: a small-target detection method. Electronics2022; 11: 1635.
28.
LiGQHaoXYZhaLL, et al. An outstanding adaptive multi-feature fusion YOLOv3 algorithm for the small target detection in remote sensing images. Pattern Anal Appl2022; 25: 951–962.
29.
WangGPYangWXLiuY, et al. Potato malformation identification and classification based on improved YOLOv3 algorithm. Electronics2023; 12: 4461.
30.
ZhuCHLiangJRZhouF. Transfer learning-based YOLOv3 model for road dense object detection. Information2023; 14: 560.
31.
WangCYLiaoHWuYH, et al. CSPNet: a new backbone that can enhance learning capability of CNN. In: Proceedings of the IEEE/CVF conference on computer vision and pattern recognition workshops, Seattle, WA, USA, 14–19 June 2020, pp.390–391. New York: IEEE.
32.
ChenXXJiangZSChengH, et al. LF-YOLOv4: a lightweight detection model for enhancing the fusion of image features of surface defects in lithium batteries. Meas Sci Technol2023; 35: 025005.
33.
YaoJQiJMZhangJ, et al. A real-time detection algorithm for Kiwifruit defects based on YOLOv5. Electronics2021; 10: 1711.
34.
QiJTLiuXNLiuK, et al. An improved YOLOv5 model based on visual attention mechanism: application to recognition of tomato virus disease. Comput Electron Agric2022; 194: 106780.
35.
HouQBZhouDQFengJS. Coordinate attention for efficient mobile network design. In: Proceedings of the IEEE/CVF conference on computer vision and pattern recognition, Nashville, TN, USA, 20–25 June 2021, pp.13713–13722. New York: IEEE.
36.
LiXTeRGYiF, et al. TCS-YOLO model for global oil storage tank inspection. Opt Precis Eng2023; 31: 246–262.
37.
LiJJinKZhouD, et al. Attention mechanism-based CNN for facial expression recognition. Neurocomputing2020; 411: 340–350.
38.
XuLSDongSHWeiHT, et al. Defect signal intelligent recognition of weld radiographs based on YOLO V5-IMPROVEMENT. J Manuf Process2023; 99: 373–381.
39.
GevorgyanZ. SIoU loss: more powerful learning for bounding box regression. arXiv preprint arXiv:220512740, 2022.
40.
CaoYKPangDDZhaoQC, et al. Improved YOLOv8-GD deep learning model for defect detection in electroluminescence images of solar photovoltaic modules. Eng Appl Artif Intell2024; 131: 107866.
41.
ZhengXLZouJXDuS, et al. Small target detection in refractive panorama surveillance based on improved YOLOv8. Sensors2024; 24: 819.
42.
GirshickRDonahueJDarrellT, et al. Rich feature hierarchies for accurate object detection and semantic segmentation. In: Proceedings of the IEEE conference on computer vision and pattern recognition, Columbus, OH, USA, 23–28 June 2014, pp.580–587. New York: IEEE.
43.
GirshickR. Fast R-CNN. In: Proceedings of the IEEE international conference on computer vision, Santiago, Chile, 7–13 December 2015, pp.1440–1448. New York: IEEE.
44.
RenSQHeKMGirshickR, et al. Faster R-CNN: towards real-time object detection with region proposal networks. Adv Neural Inf Process Syst2015; 28: 91–99.
45.
ZhangXXWangCYJinJ, et al. Object detection of VisDrone by stronger feature extraction FasterRCNN. J Electron Imaging2023; 32: 013018.
46.
ThammasanyaTPatiamSRodcharoenE, et al. A new approach to classifying polymer type of microplastics based on Faster-RCNN-FPN and spectroscopic imagery under ultraviolet light. Sci Rep2024; 14: 3529.
47.
HouJWYangCHeYH, et al. Detecting diseases in apple tree leaves using FPN–ISResNet–Faster RCNN. Eur J Remote Sens2023; 56: 2186955.
48.
LiAJChenZHNingDH, et al. Front vehicle detection based on improved fusion method for lidar and visual image. Proc IMechE, Part D: J Automobile Engineering2022; 236: 3344–3360.
49.
TangBXiangZYFanZY, et al. Effect of the perforated structure of friction block on the tribological behavior of a high-speed train brake interface in sandy environments. Eng Fail Anal2024; 158: 108039.
50.
ZhuXDChenXJ. Effect of U-shaped stress relief groove on crack initiation life of fretting fatigue. Eng Fail Anal2024; 156: 107860.
51.
YangHJGaoJXHengF, et al. A deep neural network method for LCF life prediction of metal materials with small sample experimental data. Met Mater Int2024; 30(6): 1667–1686.
52.
LiJFGuMHLiYL, et al. Predictive analysis of the precision remaining useful life of CNC machine tools based on meta-action unit. Proc IMechE, Part B: J Engineering Manufacture2025; 239(1–2): 185–197.
53.
WangLZhuSPLuoC, et al. Defect driven physics-informed neural network framework for fatigue life prediction of additively manufactured materials. Philos Trans R Soc A2023; 381: 20220386.
54.
ZhouTTJiangSHanT, et al. A physically consistent framework for fatigue life prediction using probabilistic physics-informed neural network. Int J Fatigue2023; 166: 107234.
55.
KeshunYGuangqiQYingkuiG. Optimizing prior distribution parameters for probabilistic prediction of remaining useful life using deep learning. Reliab Eng Syst Saf2024; 242: 109793.
56.
ZhengZHChenNXWuJH, et al. EW-YOLOv7: a lightweight and effective detection model for small defects in electrowetting display. Processes2023; 11: 2037.
57.
GaoWWShanMTSongN, et al. Detection of microaneurysms in fundus images based on improved YOLOv4 with SENet embedded. J Biomed Eng2022; 39: 713–720.
58.
ZhuXFLiuJYZhouXY, et al. Enhanced feature Fusion structure of YOLO v5 for detecting small defects on metal surfaces. Int J Mach Learn Cybern2023; 14: 2041–2051.
