Detection of defects in faulty elements of power lines using neural networks of YOLO

Currently, visual state diagnostics of power transmission lines (PTL) elements is a complex and time-consuming procedure. In order to increase the efficiency and reduce the labor costs of this undertaking, the most promising measure is the use of unmanned aerial vehicles equipped with computer vision systems that automatically detect damaged elements of power lines. For the purposes of improving the detection quality of power lines damaged areas by computer vision systems, the application of modern deep neural network architectures would be most effective. However, the problem of utilizing such architectures in the aforementioned task is not sufficiently covered in modern research. The issue of comparing various neural networks and identifying substantial differences in their results is especially acute. This article is devoted to a comparative analysis of modern neural network detectors YOLOv3 and YOLOv4 as well as their reduced versions (YOLOv3-tiny and YOLOv4-tiny) in terms of detecting power transmission line defects. The results of training these detectors on the CPLID dataset are presented along with statistical comparison of the YOLOv3 and YOLOv4 results by means of the cross-validation procedure. The detectors displayed high rates of detection accuracy (mAP @ 0.50 = 0.97 ± 0.03; mAP @ 0.75 = 0.78 ± 0.04) and statistically significant distinctions in these results. A comparative analysis of the findings has revealed that the employment of a simpler neural network YOLOv3 has more potential when applied to detection of power transmission line defects.

1. Savvaris A., Xie Y., Malandrakis K., Lopez M., Tsourdos A. Development of a fuel cell hybrid-powered unmanned aerial vehicle. 2016 24th mediterranean conference on control and automation (MED). 2016:1242–1247.

2. Zhang Tao, et al. Current trends in the development of intelligent unmanned autonomous systems. Frontiers of information technology & electronic engineering. 2017;18(1):68-85.

3. Sadykova D., Pernebayeva D., Bagheri M., James A. IN-YOLO: Real-time detection of outdoor high voltage insulators using UAV imaging. IEEE Transactions on Power Delivery. 2019;35(3):1599–1601.

4. Bian J., Hui X., Zhao X., Tan M. A monocular vision-based perception approach for unmanned aerial vehicle close proximity transmission tower inspection. International Journal of Advanced Robotic Systems. 2019;16(1):1729881418820227.

5. He K., Gkioxari G., Dollar P., Girshick R. Mask R-CNN. Proceedings of the IEEE international conference on computer vision. 2017:2961–2969.

6. Girshick R. Fast R-CNN. Proceedings of the IEEE international conference on computer vision. 2015:1440–1448.

7. Ren S., He K., Girshick R., Sun J. Faster r-cnn: Towards real-time object detection with region proposal networks. IEEE transactions on pattern analysis and machine intelligence. 2016;39(6):1137–1149.

8. Redmon J., Divvala S., Girshick R., Farhadi A. You only look once: Unified, real-time object detection. Proceedings of the IEEE conference on computer vision and pattern recognition. 2016:779–788.

9. Pál D., Póczos B., Szepesvári C. Estimation of Rényi Entropy and Mutual Information Based on Generalized Nearest-Neighbor Graphs. Advances in neural information processing systems. 2010:1849–1857.

10. Liu W., Anguelov D., Erhan D., Szegedy C., Reed S., Fu C.-Y., Berg A.C. SSD: Single shot multibox detector. European conference on computer vision. Springer, Cham. 2016:21–37.

11. Astapova M.A.. Lebedev I.V. Obzor podkhodov k detektirovaniyu defektov elementov LEP na izobrazheniyakh v infrakrasnom. ultrafioletovom i vidimom spektrakh. Modelirovaniye, optimizatsiya i informatsionnyye tekhnologii = Modeling, optimization and information technology. 2020;8(4):38-39. (In Russ.)

12. Prates R.M., Cruz R., Marotta A.P., Ramos R.P., Simas Filho E.F., Cardoso J.S. Insulator visual non-conformity detection in overhead power distribution lines using deep learning. Computers & Electrical Engineering. 2019;78:343–355.

13. Liu C., Wu Y., Liu J., Sun Z. Improved YOLOv3 Network for Insulator Detection in Aerial Images with Diverse Background Interference. Electronics. 2021;10(7):771.

14. Tao X., Zhang D., Wang Z., Liu X., Zhang H., Xu D. Detection of Power Line Insulator Defects Using Aerial Images Analyzed With Convolutional Neural Networks. IEEE Transactions on Systems, Man, and Cybernetics: Systems. 2020;50(4):1486–98.

15. Liao G.P., Yang G.J., Tong W.T., Gao W., Lv F.L., Gao D. Study on power line insulator defect detection via improved faster region-based convolutional neural network. 2019 IEEE 7th International Conference on Computer Science and Network Technology (ICCSNT). 2019:262–266.

16. Insulator Data Set – Chinese Power Line Insulator Dataset (CPLID). Available at: https://github.com/InsulatorData/InsulatorDataSet (accessed 16.12.2021).

17. Transmission Tower DataSet in VOC data format. Available at: https://drive.google.com/drive/folders/1UyP0fBNUqFeoW5nmPVGzyFG5IQZcqlc5 (accessed 16.12.2021).

18. Ömer Emre Yetgin, Ömer Nezih GEREK. Powerline Image Dataset (Infrared-IR and Visible Light-VL). 2019. Available at: https://data.mendeley.com/datasets/n6wrv4ry6v/8 (accessed 16.12.2021).

19. Dataset for insulator fault detection. Available at: https://figshare.com/articles/dataset/66KVimage_zip/14992944 (accessed 16.12.2021).

20. STN PLAD: A Dataset for Multi-Size Power Line Assets Detection in High-Resolution UAV Images. Available at: https://github.com/andreluizbvs/PLAD (accessed 16.12.2021).

21. Lin T.Y., Dollár P., Girshick R., He K., Hariharan B., Belongie S. Feature pyramid networks for object detection. Proceedings of the IEEE conference on computer vision and pattern recognition. 2017:2117–2125.

22. Russakovsky O., et al. Imagenet large scale visual recognition challenge. International journal of computer vision. 2015;115(3):211–252.

23. Redmon J., Farhadi A. Yolov3: An incremental improvement. arXiv preprint arXiv:1804.02767. 2018.

24. Redmon, J., Farhadi, A. Yolo9000: better, faster, stronger arXiv preprint. 2017.

25. Redmon J., et al. You only look once: Unified, real-time object detection. Proceedings of the IEEE conference on computer vision and pattern recognition. 2016:779–788.

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

27. Everingham M., Van Gool L., Williams C. K., Winn J., Zisserman A. The pascal visual object classes (voc) challenge. International journal of computer vision. 2010;88(2):303–338.

28. Lin T.Y., et al. Microsoft coco: Common objects in context. European conference on computer vision. Springer, Cham. 2014:740–755.

29. Adelson E.H., Anderson C.H., Bergen J.R., Burt P.J., Ogden J.M. Pyramid methods in image processing. RCA engineer. 1984;29(6):33–41.

30. Liu S., Qi L., Qin H., Shi J., Jia J. Path aggregation network for instance segmentation. Proceedings of the IEEE conference on computer vision and pattern recognition. 2018:8759–8768.

31. Huang G., Liu Z., Van Der Maaten L., Weinberger K.Q. Densely connected convolutional networks. Proceedings of the IEEE conference on computer vision and pattern recognition. 2017:4700–4708.

32. Bochkovskiy A., Wang C.Y., Liao, H.Y.M. Yolov4: Optimal speed and accuracy of object detection. arXiv preprint arXiv:2004.10934. 2020.

33. Dietterich T.G. Approximate statistical tests for comparing supervised classification learning algorithms. Neural computation. 1998;10(7):1895–1923.

34. Zhang K., Zhang Z., Li Z., Qiao Y. Joint face detection and alignment using multitask cascaded convolutional networks. IEEE Signal Processing Letters. 2016;23(10):1499–1503.

35. Viola P., Jones M. Rapid object detection using a boosted cascade of simple features. Proceedings of the 2001 IEEE computer society conference on computer vision and pattern recognition. CVPR 2001. 2001;1:I–I.