A Study on the Classification of Surface Defect Based on Deep Convolution Network and Transfer-learning

신경망과 전이학습 기반 표면 결함 분류에 관한 연구

  • Kim, Sung Joo (Graduate school, Korea National University of Transportation) ;
  • Kim, Gyung Bum (Aeronautical & Mechanical Design Engineering, Korea National University of Transportation)
  • 김성주 (한국교통대학교 대학원) ;
  • 김경범 (한국교통대학교 기계자동차항공공학부)
  • Received : 2021.03.09
  • Accepted : 2021.03.18
  • Published : 2021.03.31


In this paper, a method for improving the defect classification performance in low contrast, ununiformity and featureless steel plate surfaces has been studied based on deep convolution neural network and transfer-learning neural network. The steel plate surface images have low contrast, ununiformity, and featureless, so that the contrast between defect and defect-free regions are not discriminated. These characteristics make it difficult to extract the feature of the surface defect image. A classifier based on a deep convolution neural network is constructed to extract features automatically for effective classification of images with these characteristics. As results of the experiment, AlexNet-based transfer-learning classifier showed excellent classification performance of 99.43% with less than 160 seconds of training time. The proposed classification system showed excellent classification performance for low contrast, ununiformity, and featureless surface images.


  1. Neogi, N., "Review of Vision-Based Steel Surface Inspection Systems", EURASIP Journal on Image and Video Processing, Vol. 2014, No. 1, pp.1-5, 2014. https://doi.org/10.1186/1687-5281-2014-50
  2. Ghorai, S., et al, "Automatic Defect Detection on HotRolled Flat Steel Products", Institute of Electrical and Electronics Engineers Trans. on Instrumentation and Measurement, Vol. 62, NO. 3, pp.612-621, 2013.
  3. Kim, SJ., Kim, GB., "A Study on the Defect Classification of Low-contrast . Uneven . Featureless Surface Using Wavelet Transform and Support Vector Machine", Journal of Semiconductor & Display Technology, Vol. 19, No. 3, pp.1-6, 2020.
  4. Yazdchi, M. R., Mahyari, A. G., Nazeri, A., "Detection and Classification of Surface Defects of Cold Rolling Mill Steel Using Morphology and Neural Network", Institute of Electrical and Electronics Engineers, pp.1071-1076, 2008.
  5. Caleb, P., Steuer, M., "Classification of surface defects on hot rolled steel using adaptive learning methods", Institute of Electrical and Electronics Engineers, Proc. 1, pp.103-108, 2000.
  6. He, D., Xu, K., Zhou, P., "Defect detection of hot rolled steels with a new object detection framework called classification priority network", Computers & Industrial Engineering, Vol. 128, pp.290-297, 2019. https://doi.org/10.1016/j.cie.2018.12.043
  7. An, M. S., Kang, D. S., "Development of Image Analysis System Using Object Classifier basd on Deep Convolutional Neural Network", The Journal of Korean Institute of Information Technology, Vol. 14, No. 5, pp.67-73, 2016.
  8. Kim, S. J., Kim, G. B., "Optimum Illumination Mechanism for Steel Plate Inspection Using Wavelet Synthetic Image and Entropy-Based Discriminant Measure", Journal of Institute of Control, Robotics and Systems, Vol. 24, No. 9, pp.838-844, 2018. https://doi.org/10.5302/j.icros.2018.18.0108
  9. Park, J. Y., Seo, B. S., "A Study on Resilient BackPropagation Neural Network Model for Estimation of Welding Properties of Flux Cored Wire", Journal of the Korean Society of Marine Engineering, Vol. 42, No. 7, pp.531-538, 2018.
  10. Han, J. S., Kwak, K. C., "Image Classification Using Convolutional Neural Network and Extreme Learning Machine Classifier Based on ReLU Function", The Journal of Korean Institute of Information Technology, Vol. 15, No. 2, pp.15-23, 2017. https://doi.org/10.14801/jkiit.2017.15.2.15
  11. Krizhevsky, A., Sutskever, I., Hinton, G., "ImageNet classification with deep convolutional neural networks", Communications of the ACM, Vol. 60, No. 6, pp.84-90, 2017. https://doi.org/10.1145/3065386
  12. Girshick, R., Donahue, J., Darrell, T., Malik, J., "Region-Based Convolutional Networks for Accurate Object Detection and Segmentation", Institute of Electrical and Electronics Engineers Transactions on Pattern Analysis and Machine Intelligence, Vol. 38, No. 1, pp.142-158, 2016.