Detection of a casting defect tracked by deep convolution neural network

  • Jinhua Lin
  • Yu Yao
  • Lin Ma
  • Yanjie Wang


In order to relieve the problem of a false and missed detection of casting defects in X-ray detection, a robust detection method based on vision attention mechanism and deep learning of feature map is proposed. The ray images are used as input sequence, the false detection is eliminated by the intra-frame attention strategy, and the missed detection is excluded by the inter-frame deep convolution neural network (DCNN) strategy. In the intra-frame detection stage, the center-peripheral difference method is proposed to simulate the difference operation of biological vision; the suspicious defect area is directly detected according to the gradient threshold in this stage. In the inter-frame learning stage, the convolution neural network is established based on deep learning strategy to extract defect feature from a suspicious defect area; a deep learning feature vector is obtained in this stage. The similarity degree of the suspicious defect area is computed by a feature vector; a casting defect is tracked by the similarity matching of the suspicious defect in continuous frames; then, the false defects (such as noise) is excluded after defect tracking. The experimental results show that the false rate and missed rate for detection of casting defects are less than 4%, and the accuracy of the defect detection is more than 96%, which proves the robustness of the proposed method.


Casting defect Deep learning Ray image Convolution neural network 


Unable to display preview. Download preview PDF.

Unable to display preview. Download preview PDF.


  1. 1.
    Ignaszak Z, Popielarski P, Krawiec K (2007) Contribute to quantitative identification of casting defects based on computer analysis of X-ray images. Arch Foundry Eng 7(4):89–94Google Scholar
  2. 2.
    Huang Q, Wu Y, Baruch J, Jiang P, Peng Y (2009) A template model for defect simulation for evaluating nondestructive testing in X-radiography. IEEE Trans Syst Man Cybern Syst Hum 39(2):466–475CrossRefGoogle Scholar
  3. 3.
    Anand RS, Kumar P (2009) Flaw detection in radiographic weldment images using morphological watershed segmentation technique. Ndt & E Int 42(1):2–8MathSciNetCrossRefGoogle Scholar
  4. 4.
    Dubey S, Shah K (2012) Analysis of various flaws detection using segmentation techniques in weld images. Int J Adv Eng Technol 3(2):765–774Google Scholar
  5. 5.
    Manjula K, Vijayarekh K, Venkatrama B (2014) Weld flaw detection using various ultrasonic techniques: a review. J Appl Sci 14(14):1529–1535CrossRefGoogle Scholar
  6. 6.
    Mery D, Filbert D (2003) Automated flaw detection in aluminum castings based on the tracking of potential defects in a radioscopic image sequence. IEEE Trans Robot Autom 18(6):890–901CrossRefGoogle Scholar
  7. 7.
    Zhang Z, Fan X, Zhang X (2017) A fast learning-based super-resolution method for copper strip defect image. Mod Phys Lett B 31(19–21):1740042CrossRefGoogle Scholar
  8. 8.
    Zhou Z (2006) Automated defects recognition technique based on multiple radiographic images. Chin J Mech Eng 42(3):73–76CrossRefGoogle Scholar
  9. 9.
    Wang T, Chen Y, Qiao M, Snoussi H (2018) A fast and robust convolutional neural network-based defect detection model in product quality control. Int J Adv Manuf Technol 94(9–12):3465–3471CrossRefGoogle Scholar
  10. 10.
    Shao J, Shi H, Du D, Wang L, Cao H (2011) Automatic weld defect detection in real-time X-ray images based on support vector machine. Int Congress Image Signal Process 4(1):1842–1846Google Scholar
  11. 11.
    Boaretto N, Centeno TM (2017) Automated detection of welding defects in pipelines from radiographic images DWDI. Ndt & E Int 86:7–13CrossRefGoogle Scholar
  12. 12.
    Liu J, Hu YM, Wu B, Frakes DH, Wang Y (2017) A specific structuring element-based opening method for rapid geometry measurement of weld pool. Int J Adv Manuf Technol 90(5–8):1465–1477CrossRefGoogle Scholar
  13. 13.
    Krummenacher G, Cheng SO, Koller S, Kobayashi S, Buhmann JM (2017) Wheel defect detection with machine learning. IEEE Trans Intell Transp Syst 99(1):1–12Google Scholar
  14. 14.
    Lewis RW, Ransing MR, Ransing RS (2002) An approach for casting defect analysis employing finite element design optimisation, media axis transformation and neural networks. Int J Cast Metal Res 15(1):41–53CrossRefGoogle Scholar
  15. 15.
    Mery D, Jaeger T, Filbert D (2002) A review of methods for automated recognition of casting defects. Insight 44(7):428–436Google Scholar
  16. 16.
    Tae-Gyu P, Young-Cheol L, Shin-Ho L (2013) Prediction of casting defects and optimization of casting process during gravity casting of Al turbo charger valve housing. In: Marquis F. (eds) Proceedings of the 8th Pacific rim international congress on advanced materials and processingGoogle Scholar
  17. 17.
    Misimi E, Øye ER, Øystein S, Mathiassen JR (2017) Robust classification approach for segmentation of blood defects in cod fillets based on deep convolutional neural networks and support vector machines and calculation of gripper vectors for robotic processing. Comput Electron Agric 139:138–152CrossRefGoogle Scholar
  18. 18.
    Masci J, Meier U, Ciresan D, Schmidhuber J (2013) Steel defect classification with max-pooling convolutional neural networks. Proc Int Joint Conf Neural Netw 20:1–6Google Scholar
  19. 19.
    Cui X, Liu Y, Zhang Y, Wang C (2017) Tire defects classification with multi-contrast convolutional neural networks. Int J Pattern Recognit Artif Intell 32(4):1–17Google Scholar
  20. 20.
    Masci J, Meier U, Fricout G, Schmidhuber J (2015) Multi-scale pyramidal pooling network for generic steel defect classification. Int Joint Conf Neural Netw 10:1–8Google Scholar
  21. 21.
    He K, Zhang X, Ren S, Sun J (2015) Spatial pyramid pooling in deep convolutional networks for visual recognition. IEEE Trans Pattern Anal Mach Intell 37(9):1904–1916CrossRefGoogle Scholar
  22. 22.
    Liu RX, Yao MH, Wang XB (2015) Defects detection based on deep learning and transfer learning. Metall Min Ind 7:312–321Google Scholar
  23. 23.
    Yu Y, Du L, Zeng C, Zhang J (2016) Automatic localization method of small casting defect based on deep learning feature. Chin J Sci Instrum 37:1364–1370Google Scholar
  24. 24.
    Jia K, Tao D, Gao S, Xu X (2017) Improving training of deep neural networks via singular value bounding. Conf Comp Vis Pattern Recognit 2017:3994–4002Google Scholar
  25. 25.
    Rocco I, Arandjelović R, Sivic J (2017) Convolutional neural network architecture for geometric matching. Conf Comp Vis Pattern Recognit 2017:39–49Google Scholar
  26. 26.
    Kuznietsov Y, Stückler J, Leibe B (2017) Semi-supervised deep learning for monocular depth map prediction. IEEE Conf Comp Vis Pattern Recognit 2017:2215–2223Google Scholar
  27. 27.
    Mousavian A, Anguelov D, Flynn J, Kosecka J (2016) 3d bounding box estimation using deep learning and geometry. Conf Comp Vis Pattern Recognit 2017:5632–5640Google Scholar
  28. 28.
    Zheng J, Wang Q, Zhao P, Wu C (2009) Optimization of high-pressure die-casting process parameters using artificial neural network. Int J Adv Manuf Technol 44(7–8):667–674CrossRefGoogle Scholar

Copyright information

© Springer-Verlag London Ltd., part of Springer Nature 2018

Authors and Affiliations

  1. 1.Mechatronic EngineeringChangchun University of TechnologyChangchunPeople’s Republic of China
  2. 2.FAW Foundry Co., LtdChangchunPeople’s Republic of China
  3. 3.Mechatronic EngineeringChinese Academy of Sciences UniversityChangchunPeople’s Republic of China

Personalised recommendations