DOI QR코드

DOI QR Code

표면 결함 검출을 위한 데이터 확장 및 성능분석

Performance Analysis of Data Augmentation for Surface Defects Detection

  • Kim, Junbong (Dept. of Electronics Engineering, Seokyeong University) ;
  • Seo, Kisung (Dept. of Electronics Engineering, Seokyeong University)
  • 투고 : 2018.03.28
  • 심사 : 2018.04.30
  • 발행 : 2018.05.01

초록

Data augmentation is an efficient way to reduce overfitting on models and to improve a performance supplementing extra data for training. It is more important in deep learning based industrial machine vision. Because deep learning requires huge scale of learning data to learn a model, but acquisition of data can be limited in most of industrial applications. A very generic method for augmenting image data is to perform geometric transformations, such as cropping, rotating, translating and adjusting brightness of the image. The effectiveness of data augmentation in image classification has been reported, but it is rare in defect inspections. We explore and compare various basic augmenting operations for the metal surface defects. The experiments were executed for various types of defects and different CNN networks and analysed for performance improvements by the data augmentations.

키워드

참고문헌

  1. Y. Park, and I. S. Kweon "Ambiguous Surface Defect Image Classification of AMOLED Displays in Smartphones," IEEE Trans. on Industrial Informatics, Vol. 12, No. 2, pp. 597-607, 2016 https://doi.org/10.1109/TII.2016.2522191
  2. J-K. Park, N. Kwon, J-H. and D. Kang "Machine Learning-Based Imaging System for Surface Defect Inspection," International Journal of Precision Engineering and Manufacturing-Green Technology, Vol. 3, No. 3, pp. 303-310, 2016 https://doi.org/10.1007/s40684-016-0039-x
  3. H. Choi and K. Seo, "CNN Based Detection of Surface Defects for Electronic Parts," Journal of Korean Institute of Intelligent Systems, Vol. 27, No. 3, pp. 2017.
  4. H. Choi and K. Seo, "Comparison of CNN Structures for Detection of Surface Defects," The Transactions of the Korean Institute of Electrical Engineers Vol. 66, No. 7, pp. 1100 1104, 2017. https://doi.org/10.5370/KIEE.2017.66.7.1100
  5. Mandar Dixit, Roland Kwitt, Marc Niethammer, Nuno Vasconcelos, "AGA, Attribute-Guided Augmentation," arXiv:1612.02559
  6. Ken Chatfield, Karen Simonyan, Andrea Vedaldi, Andrew Zisserman, "Return of the Devil in the Details: Delving Deep into Convolutional Nets," arXiv:1405.3531
  7. Andrew G. Howard, "Some Improvements on Deep Convolutional Neural Network Based Image Classification," arXiv:1312.5402
  8. Ian J. Goodfellow, Jean Pouget-Abadie, Mehdi Mirza, Bing Xu, David Warde-Farley, Sherjil Ozair, Aaron Courville, Yoshua Bengio, "Generative Adversarial Networks," arXiv: 1406.2661
  9. Jason Wang, Luis Perez, "The Effectiveness of Data Augmentation in Image Classification using Deep Learning," Technical Report, 2017
  10. A. Krizhevsky, I. Sutskever, and G. Hinton. "ImageNet classification with deep convolutional neural networks", In NIPS, 2012.