Detection of Faults in Rail Components with Image Processing Techniques

Öz

In this study, image processing method were utilized for the early detection of defects and faults in the rails used on the railways and the components around these rails. Through early detection of the failures occurring in the railway components, this study aims to remove these failures rapidly and in an effective way; and prevent probable accidents and losses that may occur. In this context, seven particular problems have been identified from four diverse components comprised of the images that no screw situated at the junction point, clamping device rotated or extracted, rail and traverse cracks exist. From the images obtained, the features were obtained by using 5 different feature extraction methods: SIFT, SURF, GLCM, LBP and HOG. Then, using the feature vectors, Classification procedures were carried out with 10 different machine learning methods such as Decision Tree (DT), Gradient Boosting Classifier (GBC), Linear Discriminant Analysis (LDA), SVM, SVC, Logistic Regression (LR), Naive Bayes (NB), Nearest Neighbors (Knn), Neural Net (NN) and Random Forest (RF). 98% success was observed with the SVM classification method, which is one of the features extracted using HOG.

Anahtar Kelimeler

• Railways
• cracked-broken rails
• surf
• sift
• glcm
• hog
• lbp

Ray Bileşenlerinde Meydana Gelen Arızaların Görüntü İşleme Teknikleri ile Tespit Edilmesi

Öz

Bu çalışmada görüntü işleme teknikleri kullanılarak demiryollarında kullanılan ray, baglantı noktaları, travers gibi bileşenlerde oluşan, kusurların ve hataların tespiti gerçekleştirilmiştir. Demiryolu bileşenlerinde oluşan hataların erken tespiti yapılarak, bu hataların hızlı ve etkin bir şekilde ortadan kaldırılması ve bu hatalardan dolayı oluşabilecek kazaların ve kayıpların önüne geçilmesi amaçlanmıştır. Bu kapsamda demiryolu bileşenlerinden olan ray görüntülerinden bağlantı noktasında vidası olmayan, sıkma aparatı dönmüş veya çıkmış olan, ray çatlakları ve travers çatlakları bulunan görüntülerden oluşan 4 farklı bileşenden 7 farklı problem tespit edilerek çalışma gerçekleştirilmiştir. Elde edilen görüntülerden öncelikle SIFT, SURF, GLCM, LBP ve HOG olmak üzere 5 farklı öznitelik çıkarım yöntemi kullanılarak öznitelikler elde edilmiştir. Daha sonra elde edilen öznitelik vektörleri kullanılarak Decision Tree (DT), Gradient Boosting Classifier (GBC), Linear Discriminant Analysis (LDA), SVM, SVC, Logistic Regression (LR), Naive Bayes (NB), Nearest Neighbors(Knn), Neural Net (NN) ve Random Forest(RF) gibi 10 farklı makine öğrenmesi yöntemleri ile sınıflandırma işlemleri gerçekleştirilmiştir. HOG kullanılarak çıkarılan özniteliklerden SVM sınıflandırma yöntemi ile %98 oranında başarı gözlenmiştir.

Anahtar Kelimeler

• DemirYolları,
• Çatlak-Kırık Raylar
• GLCM
• HOG
• LBP
• SURF
• SIFT

Kaynakça

1. A. Lasisi, N. Attoh-Okine, “Principal components analysis and track quality index: a machine learning approach” Transp. Res. Part C: Emerg. Technol., 91, 230-248, 2018.
2. M. Bocciolone, A. Caprioli, A. Cigada, A. Collina, “A measurement system for quick rail inspection and effective track maintenance strategy”, Mechanical Systems and Signal Processing, 21(3), 1242-1254, 2007.
3. L, Zhuang, L, Wang, Z. Zhang, K.L. Tsui, “Automated vision inspection of rail surface cracks: a double-layer data-driven framework”, Transp, Res, Part C Emerg, Technol, 92, 258–277, 2018.
4. M. Chenariyan Nakhaee, D. Hiemstra, M. Stoelinga, M. van Noort, “The Recent Applications of Machine Learning in Rail Track Maintenance: A Survey”, Lecture Notes in Computer Science, 91–105, 2019.
5. I. Durazo-Cardenas, et al, “An autonomous system for maintenance scheduling data-rich complex infrastructure: fusing the railways’ condition, planning and cost”, Transp.Res.Part C Emerg.Technol, 89, 234–253 2018.
6. X. Gibert, V.M. Patel, R. Chellappa, Deep multitask learning for railway track inspection,IEEE Trans, Intell, Transp, Syst, 18, 153–164, 2017.
7. S. Öztürk, N. Öztürk, “Yapay Arı Koloni Algoritması Kullanılarak Görüntü İyileştirme Yönteminin Geliştirilmesi” Gazi Üniversitesi Fen Bilimleri Dergisi Part C: Tasarım ve Teknoloji, 4(4), 173-183, 2016.
8. R. Biswas , R. A. Khan , S. Islam , J. Uddin, “A Novel Approach to Detect and Classify the Defective of Missing Rail Anchorsin Real-time”, International Journal of Emerging Technology and Advanced Engineering, 6(12), 2016.

9. J. Liu, B. Li, Y. Xiong, B. He, L. Li, “Integrating the Symmetry Image and Improved Sparse Representation for Railway Fastener Classification and Defect Recognition”, Hindawi Publishing Corporation, Mathematical Problems in Engineering, 2015.
10. M. Karakose, O. Yaman, K. Murat, E. Akin, “A new approach for condition monitoring and detection of rail components and rail track in railway”, International Journal of Computational Intelligence Systems, 11(1), 830-845, 2018.
11. Y. Xia, F. Xie, Z. Jiang, “Broken railway fastener detection based on adaboost algorithm”, International Conference on Optoelectronics and Image Processing (ICOIP ‟10), 313–316, Haiko, China, November 2010.
12. H. Serce, Y. Bastanlar, A. Temizel, Y. Yardimci, “On Detection of Edges and Interest Points for Omnidirectional Images in Spheria Domain”, SIU 2008, Didim , Aydın, 20-22 April, 2008.
13. F. Amasyali, A. Münük, "Kamera görüntülerinden gidilen yolun kestirimi", Engineering Sciences 6(1), 296-304, 2011.
14. D. Lowe, “Object recognition from local scale-invariant features”, Proceedings of the International Conference on Computer Vision, 2, 1150–1157, 1999.
15. B. Akan, M. Çetin, A. Erçil, “Stereo based 3D head pose tracking using the scale invariant featrue transform”, 2008 IEEE 16th Signal Processing, Communication and Applications Conference, IEEE, 1-4, 2008.
16. H. Bay, A. Ess, T. Tuytelaars, L. Van Gool, “Speeded-up robust features (SURF)”, Computer vision and image understanding, 110(3), 346-359, 2008.
17. Ü. Atila, K. Akyol, F. Sabaz, “Retinal Görüntülerde Eksuda Lezyonlarının Tespiti Üzerine Bir Çalışma”, Bilişim Teknolojileri Dergisi, 13(1), 27-36, 2020.
18. Y. Aydın, “Dizdaroğlu, Bekir. Image inpainting with local feature extraction”, 2018 26th Signal Processing and Communications Applications Conference (SIU), IEEE, 1-4, 2018.
19. K. Mikolajczyk, C. Schmid, “A performance evaluation of local descriptors”, IEEE Transactions on Pattern Analysis and Machine Intelligence (TPAMI), 27(10), 1615-1630, 2005.
20. B. Yang, S. Chen, “A comparative study on local binary pattern (LBP) based face recognition: LBP histogram versus LBP image”, Neurocomputing, 365-379, 2013.
21. R. M. Haralick, K. Shanmugan “Dinstein, Its' Hak, Textural features for image classification”, IEEE Transactions on systems, man, and cybernetics, 6, 610-621, 1973.
22. S.N. Ondimu, H. Murase, Effect of probability-distance based Markovian texture extraction on discrimination in biological imaging, Computers and Electronics in Agriculture, 63(1), 2-12, 2008.
23. C.S. Hemalatha, V. Vaidehi, R. Lakshimi, “Minimal infrequent pattern based approach for mining outliers in data streams”, Expert Systems with Applications, 42(4), 1998-2012, 2015.
24. N. Dalal, B. Triggs, “Histograms of oriented gradients for human detection,” IEEE Computer Society Conference on Computer Vision and Pattern Recognition, IEEE, 1, 886–893, 2005.