One Dimensional Fourier Transform on Deep Learning for Industrial Welding Quality Control
- 1.2k Downloads
This paper presents a method for industrial welding quality control. It focuses on the detection of Lack of Fusions (LoF) in joined parts produced in a rotational welding process. The solutions are based on the LeNet and AlexNet networks that are extended with previous convolutional layers based on 1D-pDFT (1D Polar Discrete Fourier Transform) and Gabor filters. The new layers add to the network the ability to deal with the images by means of knowledge arising from the physical process. In this paper a detailed description of the optical setup and the procedure to obtain defectives samples is also given.
Keywords1D Polar Discrete Fourier Transform Deep learning AlexNet Gabor filter Industrial application Quality control
This work has been funded by the project KK-2018/00104 (Departamento de Desarrollo Económico e Infraestructuras del Gobierno Vasco, Programa ELKARTEK-Convovatoria 2018).
AM, JADB and RM developed the image analysis software and contributed mainly in the writing of the paper. XAV, MM and AGDY developed the inspection system (optics, lighting and automation). AM and JADB made the manual curation for classification by visual inspection of captured images.
- 5.Jung, S.Y., Tsai, Y.H., Chiu, W.Y., Hu, J.S., Sun, C.T.: Defect detection on randomly textured surfaces by convolutional neural networks. In: 2018 IEEE/ASME International Conference on Advanced Intelligent Mechatronics (AIM), pp. 1456–1461, July 2018Google Scholar
- 7.Kiranyaz, S., Gastli, A., Ben-Brahim, L., Alemadi, N., Gabbouj, M.: Real-time fault detection and identification for MMC using 1D convolutional neural networks. IEEE Trans. Ind. Electron. 1–1 (2018)Google Scholar
- 9.Wang, J., Yang, Y., Mao, J., Huang, Z., Huang, C., Xu, W.: CNN-RNN: a unified framework for multi-label image classification. In: The IEEE Conference on Computer Vision and Pattern Recognition (CVPR), June 2016Google Scholar
- 11.Simpson, A.J.R.: Over-sampling in a deep neural network. CoRR, vol. abs/1502.03648 (2015)Google Scholar
- 12.Ando, S., Huang, C.: Deep over-sampling framework for classifying imbalanced data. CoRR, vol. abs/1704.07515 (2017)Google Scholar
- 13.Krizhevsky, A., Sutskever, I., Hinton, G.E.: Imagenet classification with deep convolutional neural networks. In: Proceedings of the 25th International Conference on Neural Information Processing Systems, NIPS 2012 (USA), vol. 1, pp. 1097–1105. Curran Associates Inc. (2012)Google Scholar