基于明度动态感知裁剪和YOLOv8的铁路沿线异常感知检测方法

doi:10.3969/j.issn.1673-4440.2025.01.003

铁路通信信号工程技术 ›› 2025, Vol. 22 ›› Issue (1): 19-27.DOI: 10.3969/j.issn.1673-4440.2025.01.003

基于明度动态感知裁剪和YOLOv8的铁路沿线异常感知检测方法

林俊亭¹，陈权²，马赫³，彭嘉维¹，柴金川⁴

1．兰州交通大学自动化与电气工程学院，兰州 730070；
2．重庆城市职业学院信息与智能制造学院，重庆 402160；
3．中国铁路沈阳局集团有限公司长春电务段，长春 116019；
4．中国铁道科学研究院集团有限公司国家铁道试验中心，北京 100015

收稿日期:2024-07-24 修回日期:2024-11-15 出版日期:2025-01-25 发布日期:2025-01-25
作者简介:林俊亭（1978—），男，教授，博士，主要研究方向：轨道交通智能感知与自主运行技术，邮箱：linjt@lzjtu.edu.c
基金资助:
国家自然科学基金地区项目（No.52162050）

Anomaly Perception and Detection Method Along Railway Lines Based on Dynamic Value Perception and Image Cropping Along with

Lin Junting¹, Chen Quan², Ma He³, Peng Jiawei¹, Chai Jinchuan⁴

1.School of Automation and Electrical Engineering, Lanzhou Jiaotong University, Lanzhou 730070, China；
2. School of Information and Intelligent Manufacturing, Chongqing City Vocational College, Chongqing 402160, China；
3. Changchun Signalling and Communication Depot, China Railway Shenyang Group Co., Ltd., Changchun 116019, China；
4. National Railway Track Test Center, China Academy of Railway Sciences Corporation Limited, Beijing 100015, China

Received:2024-07-24 Revised:2024-11-15 Online:2025-01-25 Published:2025-01-25

摘要/Abstract

摘要： 铁路轨道沿线场景具有空旷开阔、占地面积大的特点。由于高清摄像头的普及，导致大部分入侵物体目标在图像中具有占比面积小，像素点数量少，进而降低算法对目标尤其是小目标的检测性能。针对该问题，提出基于明度动态感知裁剪和YOLOv8的铁路沿线异常感知检测方法。首先通过明度动态感知方法感知运动目标区域，再以动态目标为中心裁剪图片输入YOLOv8网络进行检测,防止目标区域图像输入网络时特征信息被压缩，最后在RailD49数据集上验证了该算法的召回率为80.33%，准确率为81.67%，相较于原始YOLOv8m对铁路运动目标的检测能力有所提升。

关键词: 目标检测, 铁路综合视频监控系统, 明度动态感知, 深度学习, YOLOv8

Abstract: Since the areas along the railway tracks are open and expansive, and as high-definition cameras are widely used, most of the intruding objects have the characteristics of occupying a small space in the images and exhibiting low pixel density, which reduces the detection performance of existing algorithms for targets, especially small ones. To address this problem, this paper proposes an anomaly perception and detection method along the railways based on dynamic value perception and image cropping, along with YOLOv8. First, the moving target area is perceived by the method of dynamic value perception. Then, the image is cropped with the dynamic target at the center and input into the YOLOv8 network for detection, to prevent the compression of the information about the key features when the target area image is fed into the network. Finally, the proposed algorithm is verified on the RailD49 dataset, with a recall of 80.33% and a precision of 81.67%, and is found to have better detection capability for moving targets along railways than original YOLOv8m.

Key words: object detection, railway integrated video surveillance system, dynamic value perception, deep learning, YOLOv8

中图分类号:

林俊亭, 陈权, 马赫, 彭嘉维, 柴金川. 基于明度动态感知裁剪和YOLOv8的铁路沿线异常感知检测方法[J]. 铁路通信信号工程技术, 2025, 22(1): 19-27.

Lin Junting, Chen Quan, Ma He, Peng Jiawei, Chai Jinchuan. Anomaly Perception and Detection Method Along Railway Lines Based on Dynamic Value Perception and Image Cropping Along with[J]. Railway Signalling & Communication Engineering, 2025, 22(1): 19-27.

图/表 10

参考文献

[1]人民铁道网.中国国家铁路集团有限公司2023年统计公报[EB/OL].(2024-03-01)[2024-05-29].https://www.peoplerail.com/rail/show-2020-532996-1.html.
[2]胡昊.面向高铁运行环境安全的侵限监测关键技术研究[D].北京:中国铁道科学研究院，2022：25-30.
[3]陈荣超.公跨营运高铁立交桥异物侵限监测系统重难点分析[J].铁路通信信号工程技术，2024，21（2）：32-36.
Chen Rongchao. Analysis of Key and Difficult Points of Foreign Object Intrusion Monitoring System for Highway Crossing High-speed Railway[J]. Railway Signalling & Communication Engineering, 2024, 21(2): 32-36.
[4]傅荟瑾.基于深度学习的高铁周界入侵监测方法研究[D].北京:中国铁道科学研究院，2023：2-3.
[5]管岭，贾利民，谢征宇.融合注意力机制的轨道入侵异物检测轻量级模型研究[J].铁道学报，2023，45（5）：72-81.
Guan Ling, Jia Limin, Xie Zhengyu. Research on Lightweight Model for Railway Intrusion Detection Integrating Attention Mechanism[J]. Journal of the China Railway Society, 2023, 45(5): 72-81.
[6]周晏.动车组受电弓状态图像检测技术研究[J].铁路通信信号工程技术，2023，20（1）：14-19.
Zhou Yan. Implementation of Image Detection System for EMU Pantograph Status[J]. Railway Signalling & Communication Engineering, 2023, 20(1): 14-19.
[7]赵宗扬，康杰虎，梁健，等.基于LGF-Net的全天候轨道入侵异物智能检测系统[J].仪器仪表学报，2023，44（9）：287-301.
Zhao Zongyang, Kang Jiehu, Liang Jian, et al. All-Weather Intelligent Detection System for Railway Intrusion Obstacles Based on LGF-Net[J]. Chinese Journal of Scientific Instrument, 2023, 44(9): 287-301.
[8]胡行涛，刘大明，虞发桐.基于FCN不确定性特征的铁路入侵异物检测算法[J].计算机应用与软件，2023，40（4）：141-146.
Hu Xingtao, Liu Daming, Yu Fatong. Foreign Intrusion Object Detection Algorithm for Railway Intrusion Based on Fcn Uncertainty[J]. Computer Applications and Software, 2023, 40(4): 141-146.
[9]赵宗扬，康杰虎，吴斌，等.基于FRL-Net的高鲁棒性多尺度小样本轨道入侵异物检测方法研究[J].仪器仪表学报，2024，45（1）：239-249.
Zhao Zongyang, Kang Jiehu, Wu Bin, et al. Research on the High Robust Multi-Scale Few-Shot Railway Intrusion Obstacles Detection Method Based on FRL-Net[J]. Chinese Journal of Scientific Instrument, 2024, 45(1): 239-249.
[10]王瑞峰，陈小屹.基于改进YOLOv5的轨道异物入侵检测算法研究[J].云南大学学报(自然科学版)，2023，45（4）：799-806.
Wang Ruifeng, Chen Xiaoyi. Research on Orbital Foreign Object Intrusion Detection Algorithm Based on Improved YOLOv5[J]. Journal of Yunnan University (Natural Sciences Edition), 2023, 45(4): 799-806.
[11]晏朋，李遇鑫，李治林，等.基于改进YOLOv7的高铁异物入侵检测算法[J].无线电工程，2024，54（5）：1099-1109.
Yan Peng, Li Yuxin, Li Zhilin, et al. High-Speed Rail Foreign Object Intrusion Detection Algorithm Based on Improved YOLOv7[J]. Radio Engineering, 2024, 54(5): 1099-1109.
[12]郑仲星，刘伟铭.基于快速扩散生成模型的地铁轨道异物入侵检测算法研究[J].铁道标准设计，2024，68（6）：191-200.
Zheng Zhongxing, Liu Weiming. Research on Foreign Object Intrusion Detection Algorithm of Metro Rail Based on Fast Diffusion Generation Model[J]. Railway Standard Design, 2024, 68(6): 191-200.
[13]胡萍.基于图像处理的城市轨道交通线路异物检测与识别[J].自动化与仪器仪表，2022（7）：86-90.
Hu Ping. Hot Spot Analysis Visualization Research Based on Big Data[J]. Automation & Instrumentation, 2022(7): 86-90.
[14] Simonyan K, Zisserman A. Very Deep Convolutional Networks for Large-Scale Iage Recognition[EB/OL]. (2014-09-04)[2024-05-29]. https://arxiv.org/abs/1409.1556.
[15] He Kaiming, Zhang Xiangyu, Ren Shaoqing, et al. Deep Residual Learning for Image Recognition[C]//2016 IEEE Conference on Computer Vision and Pattern Recognition (CVPR). June 27-30, 2016, Las Vegas, NV, USA.IEEE, 2016: 770-778.
[16] Howard A, Sandler M, Chen Bo, et al. Searching for MobileNetV3[C]//2019 IEEE/CVF International Conference on Computer Vision (ICCV).October 27-November 2, 2019, Seoul, Korea (South). IEEE, 2019: 1314-1324.
[17] Redmon J, Farhadi A. Yolov3:An Incremental Improvement[EB/OL]. (2018-04-08)[2024-05-29]. https://arxiv.org/abs/1804.02767.
[18] Jocher G, YOlOv8[EB/OL]. (2023-01-10)[2024-05-29]. https://github.com/ultralytics/ultralytics?tab=readme-ov-file.
[19] Ronneberger O, Fischer P, Brox T. U-Net: Convolutional Networks for Biomedical Image Segmentation[M]//Navab N, Hornegger J, Wells W M, et al.Lecture Notes in Computer Science. Cham: Springer International Publishing, 2015: 234-241.
[20]黄新波.基于图像感知的输电线路智能巡检综述[J].高电压技术，2024，50（5）：1826-1841.
Huang Xinbo. Survey of Intelligent Inspection Based on Image Perception[J]. High Voltage Engineering, 2024, 50(5): 1826-1841.
[21]赵红专，张继康，潘佳雯，等.基于改进视觉算法的自动驾驶风险预判模型[J].交通运输系统工程与信息，2024，24（5）：79-90，139.
Zhao Hongzhuan, Zhang Jikang, Pan Jiawen, et al. Automatic Driving Risk Prediction Model Based on Improved Vision Algorithm[J]. Journal of Transportation Systems Engineering and Information Technology, 2024, 24(5): 79-90, 139.
[22]许鲲.基于多目标识别的智能安防视频监控研究[J].自动化技术与应用，2024，43（7）：168-171.
Xu Kun. Research on Intelligent Security Video Surveillance Based on Multi-Target Recognition[J]. Techniques of Automation and Applications, 2024, 43(7): 168-171.
[23]朱力强，许力之，赵文钰，等.铁路周界入侵目标多尺度特征感知算法[J].中国铁道科学，2024，45（1）：215-226.
Zhu Liqiang, Xu Lizhi, Zhao Wenyu, et al. Multi-Scale Feature Perception Algorithm for Railway Perimeter Intrusion Object[J]. China Railway Science, 2024, 45(1): 215-226.
[24]李传，谢征宇，李永玲，等.高速铁路周界入侵视频图像样本库[J].中国铁路，2021（3）：136-143.
Li Chuan, Xie Zhengyu, Li Yongling, et al. Video Image Sample Library of HSR Perimeter Intrusion[J]. China Railway, 2021(3): 136-143.
[25]王昊.铁路综合视频监控系统技术规范发展与解析[J].中国铁路，2023（2）：113-119.
Wang Hao. Development and Analysis of Technical Specifications for Railway Integrated Video Surveillance System[J]. China Railway, 2023(2): 113-119.
[26] Lin Junting, Peng Jiawei. Adaptive Inverse Perspective Mapping Transformation Method for Ballasted Railway Based on Differential Edge Detection and Improved Perspective Mapping Model[J]. Digital Signal Processing, 2023, 135: 103944.
[27]李传.基于计算机视觉的铁路周界分割及异常感知研究[D].北京:北京交通大学，2022：3-4.
[28]孙志科.应用图像识别技术开发联锁自动测试工具的研究[J].铁路通信信号工程技术，2023，20（2）：7-12.
Sun Zhike. Research on Development of Interlocking Automatic Test Tool Using Image Recognition Technology[J]. Railway Signalling & Communication Engineering, 2023, 20(2): 7-12.

基于明度动态感知裁剪和YOLOv8的铁路沿线异常感知检测方法

Anomaly Perception and Detection Method Along Railway Lines Based on Dynamic Value Perception and Image Cropping Along with

PDF

可视化

摘要/Abstract

引用本文

使用本文

图/表 10

参考文献

相关文章 7

编辑推荐

Metrics

本文评价

[1]	高源, 张景雯, 庄文林. 面向铁路综合视频监控系统的密钥技术研究及方案应用[J]. 铁路通信信号工程技术, 2025, 22(2): 71-77.
[2]	晋云功. 基于YOLOv8算法的轨道信号灯检测研究[J]. 铁路通信信号工程技术, 2024, 21(9): 92-98.
[3]	艾博慧. 铁路综合视频一体化安全防护平台研究[J]. 铁路通信信号工程技术, 2024, 21(2): 37-42,47.
[4]	周晏. 动车组受电弓状态图像检测技术研究[J]. 铁路通信信号工程技术, 2023, 20(1): 14-19.
[5]	童伟. 一种改进的图像灰度匹配方法[J]. 铁路通信信号工程技术, 2022, 19(6): 87-90.
[6]	尹兆杰. 基于人机交互的深度学习训练数据标注系统[J]. 铁路通信信号工程技术, 2021, 18(8): 24-30.
[7]	王钏文, 王磊, 黄仁欢, 覃锐. 基于YOLOv3算法的中低速列车在途障碍物检测方法[J]. 铁路通信信号工程技术, 2021, 18(7): 86-89.