机器视觉技术在轨道交通中弯道测距的应用研究

doi:10.16180/j.cnki.issn1007-7820.2023.06.006

摘要/Abstract

摘要：

实时检测列车之间的距离是保障轨道交通安全并提高列车运能的重要途径。位于弯轨上列车之间的距离测量是测距的难点。为了推算和优化列车在弯道处的实时测距计算式,文中提出了一种基于机器视觉在轨运行列车的弯道测距方法。该方法通过单目摄像机对轨道进行图像采集,利用OpenCV算法对采集到的图片进行处理和分析,采用神经网络方法拟合弯轨轨道线,实现动态提取所需要的轨道特征点。通过搭建数学模型可以推算和优化列车之间的测距计算式。实验结果表明,在时间间距为40~50 ms时,测距系统的整体误差率小于9.11%,测距与图像中获得到的其他信息可进行有效融合与集成,且能作为提高列车安全性能的重要判据之一。

关键词: 轨道检测, 神经网络, OpenCV, 机器视觉, 弯道测距, 距离测量, 单目测距, 数学建模

Abstract:

Real-time detection of the distance between trains is an important way to ensure the safety of rail transit and improve train capacity. In particular, the distance measurement between trains on curved rails is difficult to distance measurement. In order to calculate and optimize the real-time distance measurement formula of the train at the curve, a method of distance measurement of the train at the curve based on machine vision is proposed. In this method, monocular camera is used for track image acquisition, OpenCV algorithm is used to process and analyze the acquired images, and neural network method is used to fit the track line of curved rail, realizing the dynamic extraction of track feature points. The distance formula between trains can be calculated and optimized by building a mathematical model. The experimental results show that the overall error rate of the testing system is less than 9.11% when the duration interval is 40~50 ms. The method for the spacing measurement can be integrated together with other information in the collected images for the information fusion and referred as an important condition in the rail system for a train running.

Key words: track detection, neural network, OpenCV, machine vision, curve distance, distance measurement, monocular ranging, mathematical modeling

中图分类号:

TN99

王慧敏,毕嘉桢,沈拓,张轩雄. 机器视觉技术在轨道交通中弯道测距的应用研究[J]. 电子科技, 2023, 36(6): 34-40.

WANG Huimin,BI Jiazhen,SHEN Tuo,ZHANG Xuanxiong. A Research on the Spacing Measurement Between Two Trains Located at A Curved Track by Machine Vision Technology[J]. Electronic Science and Technology, 2023, 36(6): 34-40.

图/表 13

图1

图2

图3

图4

图5

图6

图7

图8

图9

图10

图11

图12

表1

参考文献 19

[1]	洪超. 我国城市轨道交通发展趋势分析[J]. 中国市场, 2021(4):53-54.
	Hong Chao. Analysis on the development trend of urban rail transit in China[J]. China Market, 2021(4):53-54.
[2]	Zhao Y L, Li Y H, Yang L M. Research on automatic drive technology of high-speed railway rail-laying vehicles[C]. Bangkok: Proceedings of the IEEE Conference on Robotics, Automation and Mechatronics, 2006:1-5.
[3]	王伟. 一种基于特征点的高速公路弯道识别算法研究[J]. 长春工程学院学报, 2020, 3(23):105-107.
	Wang Wei. Research on a freeway curve recognition algorithm based on feature points[J]. Journal of Changchun Institute of Engineering, 2020, 3(23):105-107.
[4]	胡延平, 王乃汉, 魏振亚, 等. 一种基于卡方统计的弯道识别算法[J]. 汽车工程学报, 2018, 8(6):446-452.
	Hu Yanping, Wang Naihan, Wei Zhenya, et al. A curve recognition algorithm based on chi-square statistics[J]. Journal of Automotive Engineering, 2018, 8(6):446-452.
[5]	Muthalagua R, Bolimera A, Kalaichelvi V. Lane detection technique based on perspective transformation and histogram analysis for self-driving cars[J]. Computers and Electrical Engineering, 2020, 85(7):185-196.
[6]	王宝锋, 齐志权, 马国成, 等. 基于线性逼近的车道线弯道识别方法[J]. 北京理工大学学报, 2016, 36(5):470-474.
	Wang Baofeng, Qi Zhiquan, Ma Guocheng, et al. Lane curve recognition method based on linear approximation[J]. Transactions of Beijing Institute of Technology, 2016, 36(5):470-474.
[7]	何鹏, 高峰, 魏厚敏. 基于Catmull-Rom样条曲线的弯曲车道线检测研究[J]. 汽车工程学报, 2015, 4(6):276-281.
	He Peng, Gao Feng, Wei Houmin. Research on curved lane line detection based on Catmull-Rom spline curve[J]. Journal of Automotive Engineering, 2015, 4(6):276-281.
[8]	郭碧, 董昱. 基于分段曲线模型的铁路轨道检测算法[J]. 铁道科学与工程学报, 2017, 14(2):355-363.
	Guo Bi, Dong Yu. Railway track detection algorithm based on piecewise curve model[J]. Journal of Railway Science and Engineering, 2017, 14(2):355-363.
[9]	D'Innocente A, Carlucci F M, Colosi M, et al. Bridging between computer and robot vision through data augmentation: A case study on object recognition[C]. Shenzhen: Proceedings of the International Conference on Computer Vision Systems, 2017:384-393.
[10]	刘军, 张睿, 胡超超. 基于车道线的车辆测距方法的测距误差分析与改进[J]. 激光与光电子学进展, 2020, 57(18):297-301.
	Liu Jun, Zhang Rui, Hu Chaochao. Analysis and improvement of ranging error of vehicle ranging method based on lane line[J]. Progress in Laser and Optoelectronics, 2020, 57(18):297-301.
[11]	Eric Matthes. Python编程从入门到实践[M]. 北京: 人民邮电出版社, 2016:92-114.
	Eric Matthes. Python programming from introduction to practice[M]. Beijing: Posts & Telecom Press, 2016:92-114.
[12]	毛星云, 冷雪飞. OpenCV3编程入门[M]. 北京: 电子工业出版社, 2015:35-67.
	Mao Xingyun, Leng Xuefei. OpenCV3 programming primer[M]. Beijing: Publishing House of Electronics Industry, 2015:35-67.
[13]	Yella S, Nyberg R G, Payvar B, et al. Machine vision approach for automating vegetation detection on railway tracks[J]. Journal of Intelligent Systems, 2013, 22(2):179-196. doi: 10.1515/jisys-2013-0017
[14]	巨志勇, 张文馨, 翟春宇. 基于改进Canny算子的垃圾图像边缘检测[J]. 电子科技, 2020, 33(8):16-20.
	Ju Zhiyong, Zhang Wenxin, Zhai Chunyu. Garbage image edge detection based on improved Canny algorithm[J]. Electronics Science and Technology, 2020, 33(8):16-20.
[15]	王永超. 基于深度学习的车道线检测[D]. 秦皇岛: 燕山大学, 2018:35-48.
	Wang Yongchao. Lane line detection based on deep learning[D]. Qinhuangdao: Yanshan University, 2018: 35-48.
[16]	胡忠闯. 基于深度学习的车道线检测算法研究[D]. 杭州: 浙江大学, 2018:25-68.
	Hu Zhongchuang. Research on lane line detection algorithm based on deep learning[D]. Hangzhou: Zhejiang University, 2018:25-68.
[17]	Neven D, Brabandere B D, Georgoulis S, et al. Towards end-to-end lane detection: An instance segmentation approach[C]. Changshu: Proceedings of the Twenty-ninth IEEE Intelligent Vehicles Symposium, 2018:286-291.
[18]	杨云辉. 基于单目视觉的工件定位技术研究[J]. 电子科技, 2019, 32(12):72-75.
	Yang Yunhui. Research on workpiece positioning technology based on monocular vision[J]. Electronic science and technology, 2019, 32(12):72-75.
[19]	Guo L, Zhao Y B, Li L H, et al. Antisideslip and antirollover safety speed controller design for vehicle on curved road[J]. Mathematical Problems in Engineering, 2014, 10(1):1-12.

图片编号	实际距离 /m	测量距离 /m	测量持续时间 /m·s^-1	错误率 /%
1	40	37.56	39	5.50
2	60	56.33	42	5.17
3	80	75.11	37	4.88
4	100	93.89	42	3.90
5	120	112.67	39	3.92
6	140	131.45	40	4.86
7	160	150.22	43	5.31
8	180	169.00	29	6.17
9	200	187.78	36	6.90