Training Data Filtering for Deep Learning Applied to Inspection of Welded Joints in Oil Pipelines

• Rafael Silva Universidade Tecnológica Federal do Paraná
• Fernando Suyama Universidade Tecnológica Federal do Paraná
• Ricardo Silva Universidade Estadual de Campinas
• Myriam Delgado Federal University of Technology of Parana

DOI: https://doi.org/10.5753/eniac.2020.12170

Abstract

Inspection of weld radiographs is an essential task for preventing leaks in oil industry pipelines. Automatic inspection is important to assist specialists whose performance can be influenced by fatigue, visual acuity and experience. A challenging issue in automatic inspection is the availability of laudable images that can be used for training classifiers. Data augmentation is an alternative, however it can produce images with little or even negative impact on the training. This article presents a method for filtering images, aiming to favor the classifier relearning process, which is done by eliminating images with low performance from the training set and further retraining the segmentation model, with a positive impact for its generalization capabilities. Experimental results of the method show that data filtering accelerates the learning process and produces an improvement in the segmentation performance.

Keywords: Artificial Neural Networks, Deep Learning, Computer Vision

References

Atlas, R. M. e Hazen, T. C. (2011). Oil biodegradation and bioremediation: a tale of the two worst spills in us history.

Baniukiewicz, P. (2014). Automated defect recognition and identification in digital radiography. Journal of Nondestructive Evaluation, 33(3):327–334.

Davis, J. R. (1989). ASM Handbook: Nondestructive evaluation and quality control, volume 17. ASM International.

Duan, F., Yin, S., Song, P., Zhang, W., Zhu, C., e Yokoi, H. (2019). Automatic welding defect detection of x-ray images by using cascade adaboost with penalty term. IEEE Access, 7:125929–125938.

Fucsok, F., Muller, C., e Scharmach, M. (2002). Reliability of routine radiographic film evaluation–an extended roc study of the human factor. In 8th European Conference on Non Destructive Testing, Barcelona, June, pages 17–21.

He, K., Zhang, X., Ren, S., e Sun, J. (2016). Deep residual learning for image recognition. In Proceedings of the IEEE conference on computer vision and pattern recognition, pages 770–778.

Krizhevsky, A., Sutskever, I., e Hinton, G. E. (2012). Imagenet classification with deep convolutional neural networks. In Advances in neural information processing systems, pages 1097–1105.

Liao, T. W. (2003). Classification of welding flaw types with fuzzy expert systems. Expert Systems with Applications, 25(1):101–111.

Lin, T.-Y., Goyal, P., Girshick, R., He, K., e Dollár, P. (2017). Focal loss for dense object detection. In Proceedings of the IEEE international conference on computer vision, pages 2980–2988.

Liu, B., Zhang, X., Gao, Z., e Chen, L. (2017). Weld defect images classification with vgg16-based neural network. In International Forum on Digital TV and Wireless Multimedia Communications, pages 215–223. Springer.

Mery, D. e Arteta, C. (2017). Automatic defect recognition in x-ray testing using computer vision. In 2017 IEEE winter conference on applications of computer vision (WACV), pages 1026–1035. IEEE.

Nacereddine, N. e Tridi, M. (2005). Computer-aided shape analysis and classification of weld defects in industrial radiography based invariant attributes and neural networks. In ISPA 2005. Proceedings of the 4th International Symposium on Image and Signal Processing and Analysis, 2005., pages 88–93. IEEE.

Rana, S. et al. (2008). Facts and data on environmental risks-oil and gas drilling operations. In SPE Asia Pacific Oil and Gas Conference and Exhibition. Society of Petroleum Engineers.

Ronneberger, O., Fischer, P., e Brox, T. (2015). U-net: Convolutional networks for biomedical image segmentation. In International Conference on Medical image computing and computer-assisted intervention, pages 234–241. Springer.

Shorten, C. e Khoshgoftaar, T. M. (2019). A survey on image data augmentation for deep learning. Journal of Big Data, 6(1):60.

Simonyan, K. e Zisserman, A. (2014). Very deep convolutional networks for large-scale image recognition. arXiv preprint arXiv:1409.1556.

Szegedy, C., Liu, W., Jia, Y., Sermanet, P., Reed, S., Anguelov, D., Erhan, D., Vanhoucke, V., e Rabinovich, A. (2015). Going deeper with convolutions. In Proceedings of the IEEE conference on computer vision and pattern recognition, pages 1–9.

Wang, Y., Shi, F., e Tong, X. (2019). A welding defect identification approach in x-ray images based on deep convolutional neural networks. In International Conference on Intelligent Computing, pages 53–64. Springer.

Zapata, J., Vilar, R., e Ruiz, R. (2012). Automatic inspection system of welding radiographic images based on ann under a regularisation process. Journal of Nondestructive Evaluation, 31(1):34–45.