Geodesic
Application of neural networks for recognizing rail structural elements in magnetic and eddy current defectograms
Modelirovanie i analiz informacionnyh sistem, Tome 25 (2018) no. 6, pp. 667-679.

Voir la notice de l'article provenant de la source Math-Net.Ru

To ensure traffic safety of railway transport, non-destructive test of rails is regularly carried out by using various approaches and methods, including magnetic and eddy current flaw detection methods. An automatic analysis of large data sets (defectgrams) that come from the corresponding equipment is an actual problem. The analysis means a process of determining the presence of defective sections along with identifying structural elements of railway tracks on defectograms. This article is devoted to the problem of recognition of rail structural element images in magnetic and eddy current defectograms. Three classes of rail track structural elements are considered: 1) a bolted joint with straight or beveled connection of rails, 2) a butt weld of rails, and 3) an aluminothermic weld of rails. Images that cannot be assigned to these three classes are conditionally considered as defects and are placed in a separate fourth class. For image recognition of structural elements in defectograms a neural network is applied. The neural network is implemented by using the open library TensorFlow. To this purpose each selected (picked out) area of a defectogram is converted into a graphic image in a grayscale with size of 20$\times$39 pixels.
Keywords: nondestructive testing, magnetic and eddy current testing, rail flaw detection, automated analysis of defectograms, neural networks. 
@article{MAIS_2018_25_6_a4,
     author = {E. V. Kuzmin and O. E. Gorbunov and P. O. Plotnikov and V. A. Tyukin and V. A. Bashkin},
     title = {Application of neural networks for recognizing rail structural elements in magnetic and eddy current defectograms},
     journal = {Modelirovanie i analiz informacionnyh sistem},
     pages = {667--679},
     publisher = {mathdoc},
     volume = {25},
     number = {6},
     year = {2018},
     language = {ru},
     url = {http://geodesic.mathdoc.fr/item/MAIS_2018_25_6_a4/}
}
TY  - JOUR
AU  - E. V. Kuzmin
AU  - O. E. Gorbunov
AU  - P. O. Plotnikov
AU  - V. A. Tyukin
AU  - V. A. Bashkin
TI  - Application of neural networks for recognizing rail structural elements in magnetic and eddy current defectograms
JO  - Modelirovanie i analiz informacionnyh sistem
PY  - 2018
SP  - 667
EP  - 679
VL  - 25
IS  - 6
PB  - mathdoc
UR  - http://geodesic.mathdoc.fr/item/MAIS_2018_25_6_a4/
LA  - ru
ID  - MAIS_2018_25_6_a4
ER  - 
%0 Journal Article
%A E. V. Kuzmin
%A O. E. Gorbunov
%A P. O. Plotnikov
%A V. A. Tyukin
%A V. A. Bashkin
%T Application of neural networks for recognizing rail structural elements in magnetic and eddy current defectograms
%J Modelirovanie i analiz informacionnyh sistem
%D 2018
%P 667-679
%V 25
%N 6
%I mathdoc
%U http://geodesic.mathdoc.fr/item/MAIS_2018_25_6_a4/
%G ru
%F MAIS_2018_25_6_a4
E. V. Kuzmin; O. E. Gorbunov; P. O. Plotnikov; V. A. Tyukin; V. A. Bashkin. Application of neural networks for recognizing rail structural elements in magnetic and eddy current defectograms. Modelirovanie i analiz informacionnyh sistem, Tome 25 (2018) no. 6, pp. 667-679. http://geodesic.mathdoc.fr/item/MAIS_2018_25_6_a4/

[1] Kuzmin E. V., Gorbunov O. E., Plotnikov P. O., Tyukin V. A., “On Finding a Threshold of Useful Signals in the Analysis of Magnetic and Eddy Current Defectograms”, Modeling and Analysis of Information Systems, 24:6 (2017), 760–771 (in Russian) | MR

[2] Kuzmin E. V., Gorbunov O. E., Plotnikov P. O., Tyukin V. A., “An Efficient Algorithm for Finding a Threshold of Useful Signals in the Analysis of Magnetic and Eddy Current Defectograms”, Modeling and Analysis of Information Systems, 25:4 (2018), 382–387 (in Russian)

[3] Markov A. A., Kuznetsova E. A., Rails flaw detection. Formation and analysis of signals, v. 1, Principles, KultInformPress, St. Petersburg, 2010 (in Russian)

[4] Markov A. A., Kuznetsova E. A., Rails flaw detection. Formation and analysis of signals, v. 2, Data interpretation, Ultra Print, St. Petersburg, 2014 (in Russian)

[5] Tarabrin V. F., Zverev A. V., Gorbunov O. E., Kuzmin E. V., “About Data Filtration of the Defectogram Automatic Interpretation by Hardware and Software Complex “ASTRA””, NDT World, 64:2 (2014), 5–9 (in Russian)

[6] Simard P. Y., Steinkraus D., Platt J. C., “Best Practices for Convolutional Neural Networks Applied to Visual Document Analysis”, 7th International Conference on Document Analysis and Recognition (ICDAR-2003), 2-Volume Set (3–6 August 2003, Edinburgh, Scotland, UK), IEEE Computer Society, 2003, 958–962

[7] Goodfellow I., Bengio Y., Courville A., Deep Learning, MIT Press, 2016 | MR | Zbl

[8] Chollet F., Deep Learning with Python, Manning Publications Co., 2018

[9] TensorFlow, https://www.tensorflow.org/