DOI : https://doi.org/10.18517/ijaseit.14.5.11045

Design of Cost-Effective Steel Surface Defect Detector Based on Deep Learning

Jundong Lee (1), Hwan Seog Kim (2), Yong Wan Ju (3), Byoungwook Kim (4), Sangmin Suh (5)
(1) Department of Multimedia Engineering, University, Gangneung-Wonju National University, Wonju, Republic of Korea
(2) Department of Information and Telecommunication Engineering, Gangneung-Wonju National University, Wonju, Republic of Korea
(3) Industrial Academic Cooperation Group, Gangneung-Wonju National University, Wonju, Republic of Korea
(4) Department of Computer Science and Engineering, Gangneung-Wonju National University, Wonju, Republic of Korea
(5) Department of Information and Telecommunication Engineering, Gangneung-Wonju National University, Wonju, Republic of Korea
Fulltext View | Download
How to cite (IJASEIT) :
Lee , Jundong, et al. “Design of Cost-Effective Steel Surface Defect Detector Based on Deep Learning”. International Journal on Advanced Science, Engineering and Information Technology, vol. 14, no. 5, Oct. 2024, pp. 1786-92, doi:10.18517/ijaseit.14.5.11045.
Citation Format :
Considering the productivity and profits of manufacturers, defects in steel products must be detected very carefully. Traditionally, machine learning-based methods such as support vector machines (SVM) have been widely used for steel surface defect detection (SDD). These machine learning methods rely on expert parameters, so their performance is susceptible to these parameters. Recently, deep learning-based methods independent of these expert parameters have been developed. The most popular of these uses is transfer learning, which allows for efficient learning by applying a previously trained model to a new problem. This approach can overcome existing methods' limitations and provide more accurate and efficient SDD solutions. In order to design the transfer learning-based SDD, input images should be modified and resized, which could cause input image distortion. Moreover, the transfer learning-based models generally have many layers and weights. This paper is motivated by the following questions. 1) Is the transfer learning-based model the best method for SDD? 2) What is the smallest model without compromising performance? This paper proposes a dedicated neural network for steel surface defect detection to answer the above questions. In addition, for cheap neural networks, the initially designed neural network is gradually reduced by monitoring the performances. We achieved a maximum f1-score of 0.978 and a minimum AUC of 0.995 from the experimental results.

T. Zhang, P. Pan, J. Zhang, and X. Zhang, “Steel Surface Defect Detection Algorithm Based on Improved YOLOv8n,” Applied Sciences, vol. 14, no. 12, p. 5325, Jun. 2024, doi:10.3390/app14125325.

I. Konovalenko, P. Maruschak, and V. Brevus, “Steel Surface Defect Detection Using an Ensemble of Deep Residual Neural Networks,” Journal of Computing and Information Science in Engineering, vol. 22, no. 1, Jul. 2021, doi: 10.1115/1.4051435.

Z. H. Wei, Y. J. Zhang, X. J. Wang, J. T. Zhou, F. Q. Dou, and Y. H. Xia, "A YOLOV8-based approach for steel plate surface defect detection," Metalurgija, vol. 63, no. 1, pp. 28-30, Jan. 2024.

M. Lee and K. Seo, "Comparison of region-based CNN methods for defects detection on metal surface", Transactions of the Korean Institute of Electrical Engineers, vol. 67, no. 7, pp. 865-870, 2018.

M. Zhou, W. Lu, J. Xia, and Y. Wang, “Defect Detection in Steel Using a Hybrid Attention Network,” Sensors, vol. 23, no. 15, p. 6982, Aug. 2023, doi: 10.3390/s23156982.

C. Lv, Z. Zhang, F. Shen, F. Zhang, and H. Su, “A Fast Surface Defect Detection Method Based on Background Reconstruction,” International Journal of Precision Engineering and Manufacturing, vol. 21, no. 3, pp. 363–375, Nov. 2019, doi: 10.1007/s12541-019-00262-2.

B. Feng, J. Wu, H. Tu, J. Tang, and Y. Kang, “A Review of Magnetic Flux Leakage Nondestructive Testing,” Materials, vol. 15, no. 20, p. 7362, Oct. 2022, doi: 10.3390/ma15207362.

F. E. Yigzew, H. Kim, S. Mun, and S. Park, “Non-destructive damage detection for steel pipe scaffolds using MFL-based 3D defect visualization,” Journal of Civil Structural Health Monitoring, vol. 14, no. 2, pp. 501–509, Nov. 2023, doi: 10.1007/s13349-023-00726-0.

X. Wen, J. Shan, Y. He, and K. Song, “Steel Surface Defect Recognition: A Survey,” Coatings, vol. 13, no. 1, p. 17, Dec. 2022, doi:10.3390/coatings13010017.

R. Archana and P. S. E. Jeevaraj, “Deep learning models for digital image processing: a review,” Artificial Intelligence Review, vol. 57, no. 1, Jan. 2024, doi: 10.1007/s10462-023-10631-z.

S. Ghorai, A. Mukherjee, M. Gangadaran, and P. K. Dutta, “Automatic Defect Detection on Hot-Rolled Flat Steel Products,” IEEE Transactions on Instrumentation and Measurement, vol. 62, no. 3, pp. 612–621, Mar. 2013, doi: 10.1109/tim.2012.2218677.

R. Zaghdoudi, A. Bouguettaya, and A. Boudiaf, “Steel surface defect recognition using classifier combination,” The International Journal of Advanced Manufacturing Technology, vol. 132, no. 7–8, pp. 3489–3505, Apr. 2024, doi: 10.1007/s00170-024-13407-z.

K. Demir, M. Ay, M. Cavas, and F. Demir, “Automated steel surface defect detection and classification using a new deep learning-based approach,” Neural Computing and Applications, vol. 35, no. 11, pp. 8389–8406, Dec. 2022, doi: 10.1007/s00521-022-08112-5.

K. Liu, N. Luo, A. Li, Y. Tian, H. Sajid, and H. Chen, “A New Self-Reference Image Decomposition Algorithm for Strip Steel Surface Defect Detection,” IEEE Transactions on Instrumentation and Measurement, vol. 69, no. 7, pp. 4732–4741, Jul. 2020, doi:10.1109/tim.2019.2952706.

Y. T. Eugene Chian and J. Tian, “Surface Defect Inspection in Images Using Statistical Patches Fusion and Deeply Learned Features,” AI, vol. 2, no. 1, pp. 17–31, Jan. 2021, doi: 10.3390/ai2010002.

Y. S. Gan, S.-S. Chee, Y.-C. Huang, S.-T. Liong, and W.-C. Yau, “Automated leather defect inspection using statistical approach on image intensity,” Journal of Ambient Intelligence and Humanized Computing, vol. 12, no. 10, pp. 9269–9285, Nov. 2020, doi:10.1007/s12652-020-02631-6.

S. Son, and S. Suh, “Deep learning Based Fire and Smoke Detection Systems in Camera,” Journal of Knowledge Information Technology and Systems, vol. 17, no. 6, pp. 1251-1258, 2022.

D. Shang, F. Zhang, D. Yuan, L. Hong, H. Zheng, and F. Yang, “Deep Learning-Based Forest Fire Risk Research on Monitoring and Early Warning Algorithms,” Fire, vol. 7, no. 4, p. 151, Apr. 2024, doi:10.3390/fire7040151.

P. Dai et al., “Multi-Scale Video Flame Detection for Early Fire Warning Based on Deep Learning,” Frontiers in Energy Research, vol. 10, Mar. 2022, doi: 10.3389/fenrg.2022.848754.

H. Kwon, and S. Suh, “Automated Opthalmic Disease Diagnosis System Using Deep Learning,” Journal of Knowledge Information Technology and Systems, vol. 17, no. 6, pp. 1089-1097, 2022.

M. Iman, H. R. Arabnia, and K. Rasheed, “A Review of Deep Transfer Learning and Recent Advancements,” Technologies, vol. 11, no. 2, p. 40, Mar. 2023, doi: 10.3390/technologies11020040.

V. Chauhan, K. D. Joshi, and B. Surgenor, “Image Classification Using Deep Neural Networks: Transfer Learning and the Handling of Unknown Images,” Engineering Applications of Neural Networks, pp. 274–285, 2019, doi: 10.1007/978-3-030-20257-6_23.

K. Glynis, Z. Kapelan, M. Bakker, and R. Taormina, “Leveraging Transfer Learning in LSTM Neural Networks for Data-Efficient Burst Detection in Water Distribution Systems,” Water Resources Management, vol. 37, no. 15, pp. 5953–5972, Oct. 2023, doi: 10.1007/s11269-023-03637-3.

K. He, X. Zhang, S. Ren, and J. Sun, “Deep Residual Learning for Image Recognition,” 2016 IEEE Conference on Computer Vision and Pattern Recognition (CVPR), Jun. 2016, doi: 10.1109/cvpr.2016.90.

I. Konovalenko, P. Maruschak, V. Brevus, and O. Prentkovskis, “Recognition of Scratches and Abrasions on Metal Surfaces Using a Classifier Based on a Convolutional Neural Network,” Metals, vol. 11, no. 4, p. 549, Mar. 2021, doi: 10.3390/met11040549.

S. Wang, X. Xia, L. Ye, and B. Yang, “Automatic Detection and Classification of Steel Surface Defect Using Deep Convolutional Neural Networks,” Metals, vol. 11, no. 3, p. 388, Feb. 2021, doi:10.3390/met11030388.

Y. Li, H. Huang, Q. Xie, L. Yao, and Q. Chen, “Research on a Surface Defect Detection Algorithm Based on MobileNet-SSD,” Applied Sciences, vol. 8, no. 9, p. 1678, Sep. 2018, doi: 10.3390/app8091678.

M. Rybczak and K. Kozakiewicz, “Deep Machine Learning of MobileNet, Efficient, and Inception Models,” Algorithms, vol. 17, no. 3, p. 96, Feb. 2024, doi: 10.3390/a17030096.

O. I. Aboulola, “Improving traffic accident severity prediction using MobileNet transfer learning model and SHAP XAI technique,” PLOS ONE, vol. 19, no. 4, p. e0300640, Apr. 2024, doi:10.1371/journal.pone.0300640.

A. G. Howard et al., "MobileNets: Efficient convolutional neural networks for mobile vision applications", arXiv:1704.04861, 2017, [online] Available: http://arxiv.org/abs/1704.04861.

S. Suh, “Deep learning Based Surface Defect Detector on Metal Surface,” Journal of Knowledge Information Technology and Systems, vol. 16, no. 5, pp. 897-904, 2021.

Z. Li, H. Li, and L. Meng, “Model Compression for Deep Neural Networks: A Survey,” Computers, vol. 12, no. 3, p. 60, Mar. 2023, doi:10.3390/computers12030060.

K. Dixit, “NEU Surface Defect Database,” Northeastern University, 2023, [online] Available: https://www.kaggle.com/datasets/kaustubhdikshit/neu-surface-defect-database/

Creative Commons License

This work is licensed under a Creative Commons Attribution 4.0 International License.

Authors who publish with this journal agree to the following terms:

    1. Authors retain copyright and grant the journal right of first publication with the work simultaneously licensed under a Creative Commons Attribution License that allows others to share the work with an acknowledgement of the work's authorship and initial publication in this journal.
    2. Authors are able to enter into separate, additional contractual arrangements for the non-exclusive distribution of the journal's published version of the work (e.g., post it to an institutional repository or publish it in a book), with an acknowledgement of its initial publication in this journal.
    3. Authors are permitted and encouraged to post their work online (e.g., in institutional repositories or on their website) prior to and during the submission process, as it can lead to productive exchanges, as well as earlier and greater citation of published work (See The Effect of Open Access).