تشخیص دو پوسته بودن فولاد با شبکه‌های عصبی مصنوعی: رویکردی مبتنی بر بینایی ماشین و یادگیری عمیق

نوع مقاله : مقاله پژوهشی

نویسندگان

1 شرکت توسعه فولاد آلیاژی ایرانیان، یزد، ایران

2 واحد فناوری اطلاعات، شرکت فولاد آلیاژی ایران، یزد، ایران

10.22034/abmir.2025.23565.1159

چکیده

در این مقاله رویکردی نوین برای تشخیص خودکار عیب دوپوستگی در میلگردهای فولادی با استفاده از شبکه‌های عصبی مصنوعی و تکنیک‌های بینایی ماشین ارائه‌شده است. عیب دوپوستگی که به‌دلیل جدا شدن لایه‌ای نازک از سطح فولاد در فرآیند نورد ایجاد می‌شود، می‌تواند منجر به کاهش خواص مکانیکی و شکست محصول نهایی گردد. روش‌های سنتی بازرسی که اغلب به‌صورت دستی و زمان‌بر هستند با چالش‌هایی مانند خطای انسانی و سرعت پایین مواجه‌اند. در این پژوهش سامانه‌ای مبتنی بر بینایی ماشین پس از فرایند نورد، جهت تشخیص دوپوستگی قرار داده‌شده است. این سامانه شامل سه دوربین صنعتی با زاویه ۱۲۰ درجه و نورپردازی کنترل‌شده است که امکان تصویربرداری پیوسته و همگام با حرکت خط تولید را فراهم می‌کند. در مرحله اول از تکنیک‌های پیش‌پردازش تصویر نظیر نرمال‌سازی و قطعه‌بندی برای بهبود کیفیت تصویر استفاده‌شده است. سپس مدل‌های یادگیری عمیق، برای طبقه‌بندی و شناسایی عیب دوپوستگی آموزش داده‌شده‌اند. نتایج ارزیابی مدل‌ها بر روی داده‌های واقعی خط تولید نشان داد که مدل MobileNet با دستیابی به دقت 25/91 درصد، بهترین عملکرد را در میان مدل‌های مورد بررسی داشته است. نتایج به دست آمده بیانگر این است که سامانه پیشنهادی از قابلیت استقرار صنعتی برخوردار بوده و می‌تواند به‌عنوان جایگزینی مؤثر برای روش‌های سنتی بازرسی عمل کند.

کلیدواژه‌ها

موضوعات

عنوان مقاله [English]

Detection of Steel Bilayer Defects Using Artificial Neural Networks: A Machine Vision and Deep Learning Approach

نویسندگان [English]

  • Mostafa Majidi 1
  • Seyedamir Makinejadsanij 2
1 Iranian Alloy Steel Development Company, Yazd, Iran
2 Information Technology Department, Iran Alloy Steel Company, Yazd, Iran
چکیده [English]

n this study, a novel approach is presented for the automatic detection of delamination defects in steel rebar using artificial neural networks and computer vision techniques. The delamination defect, which occurs due to the separation of a thin surface layer of steel during the rolling process, can lead to a reduction in mechanical properties and potential failure of the final product. Traditional inspection methods, which are typically manual and time-consuming, face significant challenges such as human error and low processing speed. In this research, a machine vision–based system was implemented immediately after the rolling process to identify delamination defects. The system employs three industrial Basler cameras positioned at 120° intervals, along with controlled xenon lighting, enabling continuous and synchronized imaging with the movement of the production line. In the initial stage, image preprocessing techniques, including normalization and segmentation, were applied to enhance image quality. Subsequently, several deep learning models were trained to classify and detect delamination defects. Evaluation of the models on real production-line data demonstrated that the MobileNet model achieved the highest accuracy of 91.25%, outperforming the other models under investigation. The obtained results indicate that the proposed system possesses industrial deployment capability and can serve as an effective and reliable alternative to traditional visual inspection methods.

کلیدواژه‌ها [English]

  • Steel defect detection
  • bilayer defect
  • artificial neural networks
  • deep learning
  • machine vision

مراجع

[1]     X. Wen, S. Jvran, H. Yu and S. Kechen, "Steel Surface Defect Recognition: A Survey," Coatings, 2023.
[2]     E. Basson, "World Steel in Figures 2022," https://worldsteel.org/data/world-steel-in-figures/world-steel-in-figures-2022/, 2022.
[3]     J. Qifan and C. Li, "A Survey of Surface Defect Detection of Industrial Products Based on A Small Number of Labeled Data," arXiv, 2022.
[4]     Y. He, X. Wen and J. Xu, "A Semi-Supervised Inspection Approach of Textured Surface Defects under Limited Labeled Samples," Coatings, vol. 11, p. 12, 2022.
[5]     Y. Chen, Y. Ding, F. Zhao, E. Zhang, Z. Wu and L. Shao, "Surface Defect Detection Methods for Industrial Products: A Review," Applied Sciences, vol. 16, p. 11, 2021.
[6]     M. Chu and R. Gong, "Invariant Feature Extraction Method Based on Smoothed Local Binary Pattern for Strip Steel Surface Defect," ISIJ International, vol. 55, no. 9, pp. 1956-1962, 2015.
[7]     M. Thanh Nhat Truong and S. Kim, "Automatic image thresholding using Otsu’s method and entropy weighting scheme for surface defect detection," Soft Computing, vol. 22, no. 2, 2018.
[8]     "Distributed defect recognition on steel surfaces using an improved random forest algorithm with optimal multi-feature-set fusion," Multimedia Tools and Applications, vol. 77, no. 13, pp. 16741 - 16770, 2018.
[9]     S. H. Darwish and A. A. Halin, "Surface Defects Classification of Hot-Rolled Steel Strips Using Multi-directional Shearlet Features," Arabian Journal for Science and Engineering, vol. 44, p. 2925–2932, 2019.
[10] X. Liu, K. Xu, P. Zhou and D. Zhou, "Improved contourlet transform construction and its application to surface defect recognition of metals," Multidimensional Systems and Signal Processing, vol. 31, no. 11, 2020.
[11] M. Liu, Y. Liu, H. Hu and L. Nie, "Genetic Algorithm and Mathematical Morphology Based Binarization Method for Strip Steel Defect Image with Non-uniform Illumination," Journal of Visual Communication and Image Representation, vol. 37, 2016.
 
[12] Y.-I. Hwang, M.-K. Seo, H. G. Oh, N. Choi, G. Kim and K.-B. Kim, "Detection and Classification of Artificial Defects on Stainless Steel Plate for a Liquefied Hydrogen Storage Vessel Using Short-Time Fourier Transform of Ultrasonic Guided Waves and Linear Discriminant Analysis," Applied Sciences, vol. 12, no. 13, 2022.
[13] W. Jinjiang, F. Peilun, R. X. and Gao, "Machine vision intelligence for product defect inspection based on deep learning and Hough transform," Journal of Manufacturing Systems, vol. 51, pp. 52-60, 2019.
[14] Krizhevsky, I. Sutskever and G. E. Hinton, "ImageNet classification with deep convolutional neural networks," Association for Computing Machinery, vol. 60, no. 6, 2017.
[15] K. Simonyan and A. Zisserman, "Very Deep Convolutional Networks for Large-Scale Image Recognition," CoRR, 2014.
[16] H. Kaiming, Z. Xiangyu, R. Shaoqing and S. Jian, "Deep Residual Learning for Image Recognition," in 2016 IEEE Conference on Computer Vision and Pattern Recognition (CVPR), 2016.
[17] S. Zhang, Q. Zhang, J. Gu and L. Su, "Visual inspection of steel surface defects based on domain adaptation and adaptive convolutional neural network," Mechanical Systems and Signal Processing, vol. 15, no. 3, 2021.
[18] H. Tajmal and S. Jongwon, "Steel Surface Defect Recognition in Smart Manufacturing Using Deep Ensemble Transfer Learning-Based Techniques," Computer Modeling in Engineering & Sciences, vol. 142, no. 1, pp. 231-250, 2025.
[19] R. Joseph, D. Santosh, G. Ross and F. Ali, "You Only Look Once: Unified, Real-Time Object Detection," in 2016 IEEE Conference on Computer Vision and Pattern Recognition (CVPR), 2016.
[20] Z. Li, X. Tian, X. Liu, Y. Liu and X. Shi, "A Two-Stage Industrial Defect Detection Framework Based on Improved-YOLOv5 and Optimized-Inception-ResnetV2 Models," Applied Sciences, vol. 12, no. 2, 2022.
[21] C. Cong, L. Hoileong and C. Ming, "Steel surface defect detection method based on improved YOLOv9," Scientific Reports, 2025.
[22] S. Youkachen, M. Ruchanurucks, T. Phatrapornnant and H. Kaneko, "Defect Segmentation of Hot-rolled Steel Strip Surface by using Convolutional Auto-Encoder and Conventional Image processing," in 2019 10th International Conference of Information and Communication Technology for Embedded Systems (IC-ICTES), 2019.
[23] Y. He, K. Song, H. Dong and Y. Yan, "Semi-supervised defect classification of steel surface based on multi-training and generative adversarial network," Optics and Lasers in Engineering, vol. 122, pp. 294-302, 2020.
[24] J. Zhang, H. Su, W. Zou, X. Gong, Z. Zhang and F. Shen, "CADN:: A weakly supervised learning-based category-aware object detection network for surface defect detection," Pattern Recogn, vol. 109, pp. 0031-3203, 2021.
[25] H. Andrew G., Z. Menglong, C. Bo, K. Dmitry, W. Weijun, W. Tobias, A. Marco and A. Hartwig, "MobileNets: Efficient Convolutional Neural Networks for Mobile Vision Applications," arXiv, 2017.
پژوهش های نظری و کاربردی هوش ماشینی
دوره 3، شماره 2
پاییز و زمستان 1404
آذر 1404
صفحه 35-52

فایل ها

هم رسانی

ارجاع به این مقاله

آمار