铁路设施无人机自主巡检算法研究综述

doi:10.3778/j.issn.1002-8331.2405-0371

摘要/Abstract

摘要： 在铁路维护与安全监测领域，无人机巡检作为一种高效且能够加快发展新质生产力的方式被业界看好。现有成果多聚焦于铁路某一具体设施场景下无人机自主巡检算法的研究，综述性文章并不多见。基于此，对铁路设施无人机自主巡检算法进行系统性分析总结。归纳了当前铁路无人机自主巡检的场景和难点，分析了无人机自主巡检在各个场景下的特有问题，对近年来巡检算法研究进行总结归纳。从分析结果来看，现有研究多侧重于异物入侵和轨道扣件等缺陷的检测，而隧道和水害防治等领域的研究则相对不足。汇总了现公开的铁路相关数据集，并对评价指标从精度、速度、复杂度进行整理，对该领域未来研究方向进行展望。以期该项工作为相关领域的研究者提供宝贵的参考和指导。

关键词: 深度学习, 铁路巡检, 无人机, 目标检测, 语义分割

Abstract: In the field of railway maintenance and safety monitoring, drone inspections are seen as an efficient method that accelerates the development of new productive forces. Current research mainly focuses on autonomous inspection algorithms for specific railway facilities, with few comprehensive reviews available. This paper systematically analyzes and summarizes drone autonomous inspection algorithms for railway facilities. It first outlines the current scenarios and challenges in railway drone inspections, then analyzes the unique issues in various scenarios, summarizing recent research on inspection algorithms. The analysis shows that most studies focus on detecting foreign object intrusion and track fastener defects, while research on tunnels and flood prevention is relatively lacking. The paper also compiles publicly available railway-related datasets and organizes evaluation metrics in terms of accuracy, speed, and complexity. Finally, it discusses future research directions in this field, aiming to provide valuable reference and guidance for researchers.

Key words: deep learning, railway inspection, unmanned aerial vehicle (UAV), object detection, semantic segmentation

翟慧英, 郝汉, 李均利, 占志峰. 铁路设施无人机自主巡检算法研究综述[J]. 计算机工程与应用, 2025, 61(7): 61-80.

ZHAI Huiying, HAO Han, LI Junli, ZHAN Zhifeng. Review of Research on Unmanned Aerial Vehicle Autonomous Inspection Algorithms for Railway Facilities[J]. Computer Engineering and Applications, 2025, 61(7): 61-80.

参考文献

[1] 王华夏, 漆泰岳, 王睿. 高速铁路隧道衬砌裂缝自动化检测硬件系统研究[J]. 铁道标准设计, 2013, 57(10): 97-102.
WANG H X, QI T Y, WANG R. Research on hardware system for automatic detection of tunnel lining crack on high-speed railway[J]. Railway Standard Design, 2013, 57(10): 97-102.
[2] 程伟. 我国高铁安全监管问题现状及对策研究[J]. 中国安全科学学报, 2022, 32(S1): 34-38.
CHENG W. Research on current situation and countermeasures of high-speed railway safety supervision in China[J]. China Safety Science Journal, 2022, 32(S1): 34-38.
[3] 胡浩泽, 田昊, 于重重, 等. 改进的YOLOv4高铁接触网部件缺陷检测[J]. 计算机仿真, 2023, 40(7): 109-113.
HU H Z, TIAN H, YU C C, et al. Defect detection and recognition of key components of high-speed rail catenary based on improved YOLOv4[J]. Computer Simulation, 2023, 40(7): 109-113.
[4] 赵宇杰, 张增辉, 张向天, 等. 5G环境下无人机在高速铁路轨道巡检中的应用[J]. 中阿科技论坛(中英文), 2022(7): 141-144.
ZHAO Y J, ZHANG Z H, ZHANG X T, et al. Study on the application of UAV in high-speed railway track inspection under 5G environment[J]. China-Arab States Science and Technology Forum, 2022(7): 141-144.
[5] 姚建平, 蔡德钩, 安再展, 等. 铁路无人机巡检研究应用现状与发展趋势[J]. 铁道建筑, 2021, 61(7): 1-4.
YAO J P, CAI D G, AN Z Z, et al. Current status and development trend of research and application of railway unmanned aerial vehicle inspection[J]. Railway Engineering, 2021, 61(7): 1-4.
[6] 何家峰, 陈宏伟, 骆德汉. 深度学习实时语义分割算法研究综述[J]. 计算机工程与应用, 2023, 59(8): 13-27.
HE J F, CHEN H W, LUO D H. Review of real-time semantic segmentation algorithms for deep learning[J]. Computer Engineering and Applications, 2023, 59(8): 13-27.
[7] 程擎, 范满, 李彦冬, 等. 无人机航拍图像语义分割研究综述[J]. 计算机工程与应用, 2021, 57(19): 57-69.
CHENG Q, FAN M, LI Y D, et al. Review on semantic segmentation of UAV aerial images[J]. Computer Engineering and Applications, 2021, 57(19): 57-69.
[8] RONNEBERGER O, FISCHER P, BROX T. U-Net: convolutional networks for biomedical image segmentation[C]//Proceedings of the 18th International Conference on Medical Image Computing and Computer-Assisted Intervention. Cham: Springer, 2015: 234-241.
[9] HE K M, GKIOXARI G, DOLLáR P, et al. Mask R-CNN[C]//Proceedings of the 2017 IEEE International Conference on Computer Vision. Piscataway: IEEE, 2017: 2980-2988.
[10] CHEN L C, ZHU Y K, PAPANDREOU G, et al. Encoder-decoder with atrous separable convolution for semantic image segmentation[C]//Proceedings of the European Conference on Computer Vision. Cham: Springer, 2018: 833-851.
[11] LIU W, ANGUELOV D, ERHAN D, et al. SSD: single shot MultiBox detector[C]//Proceedings of the 14th European Conference on Computer Vision. Cham: Springer, 2016: 21-37.
[12] REDMON J, DIVVALA S, GIRSHICK R, et al. You only look once: unified, real-time object detection[C]//Proceedings of the 2016 IEEE Conference on Computer Vision and Pattern Recognition. Piscataway: IEEE, 2016: 779-788.
[13] GIRSHICK R, DONAHUE J, DARRELL T, et al. Rich feature hierarchies for accurate object detection and semantic segmentation[C]//Proceedings of the 2014 IEEE Conference on Computer Vision and Pattern Recognition. New York: ACM, 2014: 580-587.
[14] HE K M, ZHANG X Y, REN S Q, et al. Spatial pyramid pooling in deep convolutional networks for visual recognition[J]. IEEE Transactions on Pattern Analysis and Machine Intelligence, 2015, 37(9): 1904-1916.
[15] GIRSHICK R. Fast R-CNN[C]//Proceedings of the 2015 IEEE International Conference on Computer Vision. Piscataway: IEEE, 2015: 1440-1448.
[16] REN S Q, HE K M, GIRSHICK R, et al. Faster R-CNN: towards real-time object detection with region proposal networks[J]. IEEE Transactions on Pattern Analysis and Machine Intelligence, 2017, 39(6): 1137-1149.
[17] WU Y P, QIN Y, JIA L M. Research on rail surface defect detection method based on UAV images[C]//Proceedings of the 2018 Prognostics and System Health Management Conference. Piscataway: IEEE, 2018: 553-558.
[18] WU Y P, QIN Y, WANG Z P, et al. A UAV-based visual inspection method for rail surface defects[J]. Applied Sciences, 2018, 8(7): 1028.
[19] AYDIN I, SEVI M, SAHBAZ K, et al. Detection of rail defects with deep learning controlled autonomous UAV[C]//Proceedings of the 2021 International Conference on Data Analytics for Business and Industry, 2021: 500-504.
[20] BOJARCZAK P, LESIAK P. UAVs in rail damage image diagnostics supported by deep-learning networks[J]. Open Engineering, 2020, 11(1): 339-348.
[21] WU Y P, QIN Y, QIAN Y, et al. Hybrid deep learning architecture for rail surface segmentation and surface defect detection[J]. Computer-Aided Civil and Infrastructure Engineering, 2022, 37(2): 227-244.
[22] RANYAL E, JAIN K. Detection of rail fasteners from aerial images using deep convolution neural networks[C]//Proceedings of the 2020 IEEE International Geoscience and Remote Sensing Symposium. Piscataway: IEEE, 2020: 2217-2220.
[23] YU Q, LIU A, YANG X X, et al. An improved lightweight deep learning model and implementation for track fastener defect detection with unmanned aerial vehicles[J]. Electronics, 2024, 13(9): 1781.
[24] YILMAZER M, KARAKOSE M. YOLOv5 based fault detection approach in railway components[C]//Proceedings of the 2023 27th International Conference on Information Technology. Piscataway: IEEE, 2023: 1-4.
[25] ZHAO H H, WU Y P, QIN Y, et al. PB-YOLO for railway fastener detection: a higher accuracy and lightweight model with PConv and BiFPN[C]//Proceedings of the 2023 Global Reliability and Prognostics and Health Management Conference. Piscataway: IEEE, 2023: 1-7.
[26] WANG C Y, BOCHKOVSKIY A, LIAO H M. YOLOv7: trainable bag-of-freebies sets new state-of-the-art for real-time object detectors[C]//Proceedings of the 2023 IEEE/CVF Conference on Computer Vision and Pattern Recognition. Piscataway: IEEE, 2023: 7464-7475.
[27] CHEN P, WU Y P, QIN Y, et al. Rail fastener defect inspection based on UAV images: a comparative study[C]//Proceedings of the 4th International Conference on Electrical and Information Technologies for Rail Transportation. Singapore: Springer, 2020: 685-694.
[28] REDMON J, FARHADI A. YOLOv3: an incremental improvement[J]. arXiv:1804.02767, 2018.
[29] YILMAZER M, KARAKOSE M, AYDIN I, et al. Transfer learning based fault detection approach for rail components[C]//Proceedings of the 2022 26th International Conference on Information Technology. Piscataway: IEEE, 2022: 1-4.
[30] WU Y P, CHEN P, QIN Y, et al. Automatic railroad track components inspection using hybrid deep learning framework[J]. IEEE Transactions on Instrumentation and Measurement, 2023, 72: 5011415.
[31] DAI Z, WANG Z P, JIA L M, et al. An improved lightweight YOLOv5 network for defect detection of rail fasteners[C]//Proceedings of the 2022 Global Reliability and Prognostics and Health Management. Piscataway: IEEE, 2022: 1-6.
[32] YILMAZER M, KARAKOSE M. Mask R-CNN architecture based railway fastener fault detection approach[C]//Proceedings of the 2022 International Conference on Decision Aid Sciences and Applications. Piscataway: IEEE, 2022: 1363-1366.
[33] ZHANG S, CHANG Y J, WANG S H, et al. An improved lightweight YOLOv5 algorithm for detecting railway catenary hanging string[J]. IEEE Access, 2023, 11: 114061-114070.
[34] LIU J H, WANG Z P, WU Y P, et al. An improved faster R-CNN for UAV-based catenary support device inspection[J]. International Journal of Software Engineering and Knowledge Engineering, 2020, 30(7): 941-959.
[35] JIANG D Z, LIU K Y, JIA L M, et al. Automatic detection strategy of multi-scale catenary support device based on improved YOLOv7[J]. IFAC-PapersOnLine, 2023, 56(2): 7597-7602.
[36] DING X W, CAI X N, ZHANG Z Y, et al. Railway foreign object intrusion detection based on deep learning[C]//Proceedings of the 2022 International Conference on Computer Engineering and Artificial Intelligence. Piscataway: IEEE, 2022: 735-739.
[37] CAI X N, DING X W. A comparative study of machine vision-based rail foreign object intrusion detection models[C]//Proceedings of the 2023 IEEE 3rd International Conference on Power, Electronics and Computer Applications. Piscataway: IEEE, 2023: 1304-1308.
[38] HUANG H R, ZHAO G P, BO Y M, et al. Railway intrusion detection based on refined spatial and temporal features for UAV surveillance scene[J]. Measurement, 2023, 211: 112602.
[39] LI Y D, DONG H, LI H G, et al. Multi-block SSD based on small object detection for UAV railway scene surveillance[J]. Chinese Journal of Aeronautics, 2020, 33(6): 1747-1755.
[40] TONG L, WANG Z P, QIN Y, et al. Leveraging UAVs for rapid and hierarchical railway intrusion detection[C]//Proceedings of the 6th International Conference on Electrical Engineering and Information Technologies for Rail Transportation. Singapore: Springer, 2024: 127-134.
[41] WU Y P, MENG F T, QIN Y, et al. UAV imagery based potential safety hazard evaluation for high-speed railroad using Real-time instance segmentation[J]. Advanced Engineering Informatics, 2023, 55: 101819.
[42] WANG S H, WANG Y D, CHANG Y J, et al. EBSE-YOLO: high precision recognition algorithm for small target foreign object detection[J]. IEEE Access, 2023, 11: 57951-57964.
[43] 李树杰, 贾子彦. 基于无人机机器视觉的轨道异物入侵检测方法[J]. 无线互联科技, 2023, 20(7): 112-117.
LI S J, JIA Z Y. Orbital foreign body intrusion detection method based on UAV machine vision[J]. Wireless Internet Technology, 2023, 20(7): 112-117.
[44] QIAO X R, LI J Q, PENG Y H, et al. Foreign body Intrusion detection based on YOLOv5 with self-attentional feature embedding[C]//Proceedings of the 2023 IEEE 7th Information Technology and Mechatronics Engineering Conference. Piscataway: IEEE, 2023: 519-524.
[45] ZHANG Z H, MENG W W, LI Z, et al. A small-scale pedestrian detection method in railway scene[C]//Proceedings of the 2022 IEEE 5th International Conference on Electronics and Communication Engineering. Piscataway: IEEE, 2022: 188-192.
[46] 侯涛, 宝才文, 陈燕楠. 基于自适应高斯混合模型的铁轨异物入侵检测研究[J]. 光电子·激光, 2022, 33(4): 403-413.
HOU T, BAO C W, CHEN Y N. Research on detection of foreign object intrusion in railroad tracks based on AGMM[J]. Journal of Optoelectronics·Laser, 2022, 33(4): 403-413.
[47] GWON G H, LEE J H, KIM I H, et al. CNN-based image quality classification considering quality degradation in bridge inspection using an unmanned aerial vehicle[J]. IEEE Access, 2023, 11: 22096-22113.
[48] WANG F, ZOU Y, CHEN X Y, et al. Rapid in-flight image quality check for UAV-enabled bridge inspection[J]. ISPRS Journal of Photogrammetry and Remote Sensing, 2024, 212: 230-250.
[49] 焦胜. 无人机在公路桥梁检测中的应用研究[J]. 全面腐蚀控制, 2024, 38(3): 30-33.
JIAO S. Research on application of UAV in highway bridge detection[J]. Total Corrosion Control, 2024, 38(3): 30-33.
[50] JIANG T J, FR?SETH G T, R?NNQUIST A, et al. A visual inspection and diagnosis system for bridge rivets based on a convolutional neural network[J]. Computer-Aided Civil and Infrastructure Engineering, 2024, 39(24): 3786-3804.
[51] MU Z H, QIN Y, YU C C, et al. Adaptive cropping shallow attention network for defect detection of bridge girder steel using unmanned aerial vehicle images[J]. Journal of Zhejiang University: Science A, 2023, 24(3): 243-256.
[52] LEE J H, GWON G H, KIM I H, et al. A motion deblurring network for enhancing UAV image quality in bridge inspection[J]. Drones, 2023, 7(11): 657.
[53] GOODFELLOW I, POUGET-ABADIE J, MIRZA M, et al. Generative adversarial networks[J]. Communications of the ACM, 2020, 63(11): 139-144.
[54] JIANG S, ZHANG J, WANG W G, et al. Automatic inspection of bridge bolts using unmanned aerial vision and adaptive scale unification-based deep learning[J]. Remote Sensing, 2023, 15(2): 328.
[55] MU Z H, QIN Y, YU C C, et al. UAV image defect detection method for steel structure of high-speed railway bridge girder[C]//Proceedings of the 2022 Global Reliability and Prognostics and Health Management. Piscataway: IEEE, 2022: 1-8.
[56] KAO S P, CHANG Y C, WANG F L. Combining the YOLOv4 deep learning model with UAV imagery processing technology in the extraction and quantization of cracks in bridges[J]. Sensors, 2023, 23(5): 2572.
[57] BOCHKOVSKIY A, WANG C Y, LIAO H M. YOLOv4: optimal speed and accuracy of object detection[J]. arXiv:2004.10934, 2020.
[58] 马学志, 范剑雄, 柴雪松, 等. 无人机巡检系统在铁路混凝土桥梁检测中的应用[J]. 铁道建筑, 2021, 61(12): 76-80.
MA X Z, FAN J X, CHAI X S, et al. Application of UAV inspection system in railway concrete bridge inspection[J]. Railway Engineering, 2021, 61(12): 76-80.
[59] ZHAO H S, SHI J P, QI X J, et al. Pyramid scene parsing network[C]//Proceedings of the 2017 IEEE Conference on Computer Vision and Pattern Recognition. Piscataway: IEEE, 2017: 6230-6239.
[60] 余加勇, 刘宝麟, 尹东, 等. 基于YOLOv5和U-Net3+的桥梁裂缝智能识别与测量[J]. 湖南大学学报(自然科学版), 2023, 50(5): 65-73.
YU J Y, LIU B L, YIN D, et al. Intelligent identification and measurement of bridge cracks based on YOLOv5 and U-Net3+[J]. Journal of Hunan University(Natural Sciences), 2023, 50(5): 65-73.
[61] HUANG H M, LIN L F, TONG R F, et al. UNet 3+: a full-scale connected UNet for medical image segmentation[C]//Proceedings of the 2020 IEEE International Conference on Acoustics, Speech and Signal Processing, 2020: 1055-1059.
[62] ZHANG R, OUYANG A H, LI Z L. Automatic UAV inspection of tunnel infrastructure in GPS-denied underground environment[C]//European Workshop on Structural Health Monitoring. Cham: Springer, 2022: 519-526.
[63] 王可心. 基于轻量注意力改进的无人机载雷达图像YOLO识别方法研究[J]. 铁道建筑技术, 2023(7): 5-8.
WANG K X. Improved YOLO recognition method for unmanned aerial vehicle radar images based on lightweight attention[J]. Railway Construction Technology, 2023(7): 5-8.
[64] LIU K, WANG M, ZHOU T J. Increasing costs to Chinese railway infrastructure by extreme precipitation in a warmer world[J]. Transportation Research Part D: Transport and Environment, 2021, 93: 102797.
[65] PHAM M V, KIM Y T. Debris flow detection and velocity estimation using deep convolutional neural network and image processing[J]. Landslides, 2022, 19(10): 2473-2488.
[66] 宗慧琳, 袁希平, 甘淑, 等. 一种优化的泥石流无人机影像匹配算法[J]. 测绘科学, 2022, 47(11): 104-112.
ZONG H L, YUAN X P, GAN S, et al. An optimized UAV image matching algorithm in debris flow areas[J]. Science of Surveying and Mapping, 2022, 47(11): 104-112.
[67] CHEN J M, XIE Z Y, JIA L M, et al. Research on railway geological hazard detection method based on few-shot deep learning[C]//Proceedings of the 5th International Conference on Electrical Engineering and Information Technologies for Rail Transportation. Singapore: Springer, 2022: 166-175.
[68] QIU H J, SU L L, TANG B Z, et al. The effect of location and geometric properties of landslides caused by rainstorms and earthquakes[J]. Earth Surface Processes and Landforms, 2024, 49(7): 2067-2079.
[69] LV Z Y, YANG T, LEI T, et al. Spatial-spectral similarity based on adaptive region for landslide inventory mapping with remote-sensed images[J]. IEEE Transactions on Geoscience and Remote Sensing, 2024, 62: 4405111.
[70] LV P Y, MA L S, LI Q M, et al. ShapeFormer: a shape-enhanced vision transformer model for optical remote sensing image landslide detection[J]. IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing, 2023, 16: 2681-2689.
[71] CHEN S J, XIANG C C, KANG Q, et al. Accurate landslide detection leveraging UAV-based aerial remote sensing[J]. IET Communications, 2020, 14(15): 2434-2441.
[72] WU Z B, LI H, YUAN S X, et al. Mask R-CNN-based landslide hazard identification for 22. 6 extreme rainfall induced landslides in the Beijiang River Basin, China[J]. Remote Sensing, 2023, 15(20): 4898.
[73] YUN L, ZHANG X X, ZHENG Y C, et al. Enhance the accuracy of landslide detection in UAV images using an improved mask R-CNN model: a case study of Sanming, China[J]. Sensors, 2023, 23(9): 4287
[74] WEI R L, YE C M, SUI T B, et al. A feature enhancement framework for landslide detection[J]. International Journal of Applied Earth Observation and Geoinformation, 2023, 124: 103521.
[75] XIE Y K, ZHAN N, ZHU J, et al. Landslide extraction from aerial imagery considering context association characteristics[J]. International Journal of Applied Earth Observation and Geoinformation, 2024, 131: 103950.
[76] NIU C Y, OUYANG G, LU W J, et al. Reg-SA-UNet++: a lightweight landslide detection network based on single-temporal images captured postlandslide[J]. IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing, 2022, 15: 9746-9759.
[77] ZHOU Z W, RAHMAN SIDDIQUEE M M, TAJBAKHSH N, et al. UNet++: a nested U-Net architecture for medical image segmentation[M]//Deep learning in medical image analysis and multimodal learning for clinical decision support. Cham: Springer, 2018: 3-11.
[78] FAN H, COSMAN P C, HOU Y, et al. High-speed railway fastener detection based on a line local binary pattern[J]. IEEE Signal Processing Letters, 2018, 25(6): 788-792.
[79] ZHANG Z H, YU S Z, YANG S W, et al. Rail-5k: a real-world dataset for rail surface defects detection[J]. arXiv: 2106.14366, 2021.
[80] GAN J R, LI Q Y, WANG J Z, et al. A hierarchical extractor-based visual rail surface inspection system[J]. IEEE Sensors Journal, 2017, 17(23): 7935-7944.
[81] GUO F, QIAN Y, WU Y P, et al. Automatic railroad track components inspection using real-time instance segmentation[J]. Computer-Aided Civil and Infrastructure Engineering, 2021, 36(3): 362-377.
[82] CHEN Z C, YANG J, FENG Z C, et al. RailFOD23: a dataset for foreign object detection on railroad transmission lines[J]. Scientific Data, 2024, 11(1): 72.
[83] REN Y P, HUANG J S, HONG Z Y, et al. Image-based concrete crack detection in tunnels using deep fully convolutional networks[J]. Construction and Building Materials, 2020, 234: 117367.