Abstract
Rod insulators are vital parts of the catenary of high speed railways (HSRs). There are many different catenary insulators, and the background of the insulator image is complicated. It is difficult to recognise insulators and detect defects automatically. In this paper, we propose a catenary intelligent defect detection algorithm based on Mask region-convolutional neural network (R-CNN) and an image processing model. Vertical projection technology is used to achieve single shed positioning and precise cutting of the insulator. Gradient, texture, and gray feature fusion (GTGFF) and a K-means clustering analysis model (KCAM) are proposed to detect broken insulators, dirt, foreign bodies, and flashover. Using this model, insulator recognition and defect detection can achieve a high recall rate and accuracy, and generalized defect detection. The algorithm is tested and verified on a dataset of realistic insulator images, and the accuracy and reliability of the algorithm satisfy current requirements for HSR catenary automatic inspection and intelligent maintenance.
Abstract
目的
绝缘子是高速铁路接触网的重要组成部分。绝缘子的故障会导致绝缘劣化, 甚至会导致接触网断电, 所以绝缘子缺陷检测对高速列车运行具有重要意义。本文旨在分析巡检车拍摄的接触网绝缘子的图像特点, 结合绝缘子破损、污垢、异物和闪络四类主要缺陷, 研究一种智能图像处理方法, 以期有效识别绝缘子及其缺陷。
创新点
1. 通过Mask R-CNN模型, 实现绝缘子区域像素级切割及旋正; 2. 提出垂直投影技术, 实现绝缘子单片区域快速准确定位; 3. 通过多特征融合和聚类分析模型, 检测绝缘子破损、污垢、异物和闪络。
方法
1. 通过分析接触网图像的特点, 采用Mask R-CNN方法实现绝缘子区域定位、前后景像素分割以及倾斜修正(图5); 2. 通过垂直投影方法, 得到绝缘子各片空间坐标信息(图6); 3. 通过提取图像梯度、纹理和灰度特征(公式(2)~(4)), 运用特征融合聚类方法(公式(5)~(7)), 计算其相邻片之间的特征分布差异(公式(8)); 4. 基于实际拍摄图片构建实验测试样本, 并分析实验过程及结果, 验证所提方法的可行性和有效性。
结论
1. Mask R-CNN 是一种高效的目标识别和实例分割深度学习模型; 它在绝缘子识别方面展现了鲁棒性和高精度。2. 实验表明, 本文提出的绝缘子像素区域切割和倾斜校正具有较高精度。3. 对于绝缘子缺陷检测, 本文提出的多特征融合聚类分析模型测试结果显示其具有较高的缺陷识别精确度。
This is a preview of subscription content, log in via an institution to check access.
Similar content being viewed by others
References
Adou MW, Xu HR, Chen GH, 2019. Insulator faults detection based on deep learning. Proceedings of the IEEE 13th International Conference on Anti-counterfeiting, Security, and Identification, p.173–177. https://doi.org/10.1109/ICASID.2019.8925094
Dai JF, He KM, Sun J, 2016. Instance-aware semantic segmentation via multi-task network cascades. Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition, p.3150–3158. https://doi.org/10.1109/CVPR.2016.343
Dalal N, Triggs B, 2005. Histograms of oriented gradients for human detection. Proceedings of the IEEE Computer Society Conference on Computer Vision and Pattern Recognition, p.886–893. https://doi.org/10.1109/CVPR.2005.177
Han Y, Liu ZG, Lee DJ, et al., 2016. High-speed railway rodinsulator detection using segment clustering and deformable part models. Proceedings of the IEEE International Conference on Image Processing, p.3852–3856. https://doi.org/10.1109/ICIP.2016.7533081
Hartigan JA, Wong MA, 1979. Algorithm as 136: a K-means clustering algorithm. Journal of the Royal Statistical Society. Series C (Applied Statistics), 28(1):100–108. https://doi.org/10.2307/2346830
He KM, Gkioxari G, Dollár P, et al., 2017. Mask R-CNN. Proceedings of the IEEE International Conference on Computer Vision, p.2980–2988. https://doi.org/10.1109/ICCV.2017.322
Jiang H, Qiu XJ, Chen J, et al., 2019. Insulator fault detection in aerial images based on ensemble learning with multi-level perception. IEEE Access, 7:61797–61810. https://doi.org/10.1109/ACCESS.2019.2915985
Kang GQ, Gao SB, Yu L, et al., 2019. Deep architecture for high-speed railway insulator surface defect detection: denoising autoencoder with multitask learning. IEEE Transactions on Instrumentation and Measurement, 68(8):2679–2690. https://doi.org/10.1109/TIM.2018.2868490
Kien DN, Zhuang XY, 2021. A deep neural network-based algorithm for solving structural optimization. Journal of Zhejiang University-SCIENCE A (Applied Physics & Engineering), 22(8):609–620. https://doi.org/10.1631/jzus.A2000380
LeCun Y, Bengio Y, Hinton G, 2015. Deep learning. Nature, 521(7553):436–444. https://doi.org/10.1038/nature14539
Li Y, Qi HZ, Dai JF, et al., 2017. Fully convolutional instance-aware semantic segmentation. Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition, p.4438–4446. https://doi.org/10.1109/CVPR.2017.472
Liu T, Zhao Y, Wei YC, et al., 2019. Concealed object detection for activate millimeter wave image. IEEE Transactions on Industrial Electronics, 66(12):9909–9917. https://doi.org/10.1109/TIE.2019.2893843
Lyu Y, Han ZW, Zhong JP, et al., 2020. A generic anomaly detection of catenary support components based on generative adversarial networks. IEEE Transactions on Instrumentation and Measurement, 69(5):2439–2448. https://doi.org/10.1109/TIM.2019.2954757
Nikbakht S, Anitescu C, Rabczuk T, 2021. Optimizing the neural network hyperparameters utilizing genetic algorithm. Journal of Zhejiang University-SCIENCE A (Applied Physics & Engineering), 22(6):407–426. https://doi.org/10.1631/jzus.A2000384
Ojala T, Pietikäinen M, Harwood D, 1996. A comparative study of texture measures with classification based on featured distributions. Pattern Recognition, 29(1):51–59. https://doi.org/10.1016/0031-3203(95)00067-4
Ren SQ, He KM, Girshick R, et al., 2017. Faster R-CNN: towards real-time object detection with region proposal networks. IEEE Transactions on Pattern Analysis and Machine Intelligence, 39(6):1137–1149. https://doi.org/10.1109/TPAMI.2016.2577031
Su M, Peng H, Li SF, 2021. Application of an interpretable artificial neural network to predict the interface strength of a near-surface mounted fiber-reinforced polymer to concrete joint. Journal of Zhejiang University-SCIENCE A (Applied Physics & Engineering), 22(6):427–440. https://doi.org/10.1631/jzus.A2000245
Tan P, Ma JE, Zhou J, et al., 2016. Sustainability development strategy of China’s high speed rail. Journal of Zhejiang University-SCIENCE A (Applied Physics & Engineering), 17(12):923–932. https://doi.org/10.1631/jzus.A1600747
Tan P, Li XF, Xu JM, et al., 2020. Catenary insulator defect detection based on contour features and gray similarity matching. Journal of Zhejiang University-SCIENCE A (Applied Physics & Engineering), 21(1):64–73. https://doi.org/10.1631/jzus.A1900341
Tan P, Li XF, Wu ZG, et al., 2021. Multialgorithm fusion image processing for high speed railway dropper failure-defect detection. IEEE Transactions on Systems, Man, and Cybernetics: Systems, 51(7):4466–4478. https://doi.org/10.1109/TSMC.2019.2938684
Wang J, Wang TT, Luo QZ, 2019. A practical structural health monitoring system for high-speed train car-body. IEEE Access, 7:168316–168326. https://doi.org/10.1109/ACCESS.2019.2954680
Wang L, Zhang ZJ, 2017. Automatic detection of wind turbine blade surface cracks based on UAV-taken images. IEEE Transactions on Industrial Electronics, 64(9):7293–7303. https://doi.org/10.1109/TIE.2017.2682037
Yao XT, Zhang YP, Gao Y, et al., 2017. Research on technology of contact insulator pollution classification and pollution level division based on machine vision. Proceedings of the IEEE 3rd Information Technology and Mechatronics Engineering Conference, p.1264–1268. https://doi.org/10.1109/ITOEC.2017.8122559
Yao XT, Zhang YP, Gao Y, et al., 2018. Research on vision-based technology of contamination detection in overhead line insulators. Journal of Lanzhou Jiaotong University, 37(1):65–72 (in Chinese). https://doi.org/10.3969/j.issn.1001-4373.2018.01.011
Yue GH, Hou CP, Gu K, et al., 2018. Biologically inspired blind quality assessment of tone-mapped images. IEEE Transactions on Industrial Electronics, 65(3):2525–2536. https://doi.org/10.1109/TIE.2017.2739708
Zeng X, Li ZH, Gao W, et al., 2018. A novel virtual sensing with artificial neural network and K-means clustering for IGBT current measuring. IEEE Transactions on Industrial Electronics, 65(9):7343–7352. https://doi.org/10.1109/TIE.2018.2793196
Zhao ZB, Zhen Z, Zhang L, et al., 2019. Insulator detection method in inspection image based on improved faster R-CNN. Energies, 12(7):1204. https://doi.org/10.3390/en12071204
Zhu Q, Li L, Chen CJ, et al., 2018. A low-cost lateral active suspension system of the high-speed train for ride quality based on the resonant control method. IEEE Transactions on Industrial Electronics, 65(5):4187–4196. https://doi.org/10.1109/TIE.2017.2767547
Acknowledgments
This work is supported by the National Natural Science Foundation of China (Nos. 51677171, 51637009, 51577166, and 51827810), the National Key R&D Program of China (No. 2018YFB0606000), the China Scholarship Council (No. 201708330502), the Fund of Shuohuang Railway Development Limited Liability Company (No. SHTL-2020-13), and the Fund of State Key Laboratory of Industrial Control Technology (No. ICT2022B29), China.
Author information
Authors and Affiliations
Contributions
Ping TAN and Xu-feng LI designed the research. Ping TAN, Xu-feng LI, Jin DING, and Zhi-sheng CUI processed the corresponding data. Ping TAN and Xu-feng LI wrote the first draft of the manuscript. Jin DING, Zhi-sheng CUI, Ji-en MA, Yue-lan SUN, Bing-qiang HUANG, and You-tong FANG helped to organize the manuscript. Ping TAN, Xu-feng LI, and Zhi-sheng CUI revised and edited the final version.
Corresponding authors
Additional information
Conflict of interest
Ping TAN, Xu-feng LI, Jin DING, Zhi-sheng CUI, Ji-en MA, Yue-lan SUN, Bing-qiang HUANG, and You-tong FANG declare that they have no conflict of interest.
Rights and permissions
About this article
Cite this article
Tan, P., Li, Xf., Ding, J. et al. Mask R-CNN and multifeature clustering model for catenary insulator recognition and defect detection. J. Zhejiang Univ. Sci. A 23, 745–756 (2022). https://doi.org/10.1631/jzus.A2100494
Received:
Accepted:
Published:
Issue Date:
DOI: https://doi.org/10.1631/jzus.A2100494
Key words
- High speed railway (HSR) catenary insulator
- Mask region-convolutional neural network (R-CNN)
- Multifeature fusion
- K-means clustering analysis model (KCAM)
- Defect detection