基于视频深度学习的铁路周界入侵检测算法研究

交通运输系统工程与信息 ›› 2020, Vol. 20 ›› Issue (2): 61-68.

• 智能交通系统与信息技术 • 上一篇下一篇

基于视频深度学习的铁路周界入侵检测算法研究

王瑞¹，李霄峰²，史天运^{* 3}，邹琪²

1. 中国铁道科学研究院研究生部，北京 100081；2. 北京交通大学轨道交通与数据挖掘北京市重点实验室，北京 100044；3. 中国铁道科学研究院集团有限公司，北京 100081

收稿日期:2019-10-23 修回日期:2019-12-19 出版日期:2020-04-25 发布日期:2020-04-30
作者简介:王瑞(1985-)，男，山西太原人，副研究员，博士生.
基金资助:
中国国家铁路集团有限公司科技研究开发计划/Science and Technology Research and Development Program of China National Railway Group Co., Ltd.(K2018T002，SY2017-T001).

Railway Perimeter Intrusion Detection Algorithms Based on Video Deep Learning

WANG Rui¹, LI Xiao-feng², SHI Tian-yun³, ZOU Qi²

1. Graduate School of China Academy of Railway Science, Beijing 100081, China; 2. Beijing Key Laboratory of Rail Transit and Data Mining, Beijing Jiaotong University, Beijing 100044, China; 3. China Academy of Railway Sciences Corporation Limited, Beijing 100081, China

Received:2019-10-23 Revised:2019-12-19 Online:2020-04-25 Published:2020-04-30

摘要/Abstract

摘要：

基于视频智能分析的铁路周界入侵检测算法相比于雷达、振动光纤，具有成本低、误报率低的优点. 针对视频中存在不同分辨率目标的问题，提出一种改进的Cascade Mask RCNN( CMR)模型，使用级联结构获得目标的准确定位. 为增强模型对小目标的检测能力，在原始模型的基础上，增加基于特征金字塔网络(FPN)的多尺度特征提取模块和基于空洞金字塔汇聚(ASPP)子网络的空间上下文增强模块. 在实际铁路周界入侵场景视频中验证了模型的有效性. 结果表明，该模型可实现不同场景下的铁路周界入侵检测，相较于原始模型，新模型对小目标检测的F-measure 提高了0.24. 模型既解决了不同场景下铁路周界入侵检测问题，又有效地提高了视频智能分析对小目标检测的准确率.

关键词: 铁路运输, 视频智能分析, 深度学习, 周界入侵检测, 特征金字塔, 空洞卷积

Abstract:

Compared with radar and vibration optical fiber, the railway perimeter intrusion detection algorithm based on video intelligent analysis has the advantages of low cost and low false alarm rate. Aiming at the simultaneous detection of large and small targets in videos, this paper proposes an improved Cascade Mask RCNN (CMR) model. The model uses the cascaded structure to locate the target accurately. At the same time, a multi- scale feature extraction model is added to the original model to enhance the expressive ability for small targets. Multi-scale module is implemented by feature pyramid networks (FPN). The spatial context enhancement module is implemented through an atrous spatial pyramid pooling (ASPP) subnetwork. The effectiveness of the proposed model was verified by the actual railway scene videos. The results show that the new model improves the F-measure of small targets detection by up to 0.24 compared with the original model. The proposed model enhances the detection capability of railway perimeter intrusion in different scenarios, and improves the accuracy of video detection for small targets.

Key words: railway transportation, video intelligent analysis, deep learning, perimeter intrusion detection, feature pyramid networks, atrous spatial pyramid pooling

中图分类号:

U298.1

王瑞，李霄峰，史天运，邹琪. 基于视频深度学习的铁路周界入侵检测算法研究[J]. 交通运输系统工程与信息, 2020, 20(2): 61-68.

WANG Rui, LI Xiao-feng, SHI Tian-yun, ZOU Qi. Railway Perimeter Intrusion Detection Algorithms Based on Video Deep Learning[J]. Journal of Transportation Systems Engineering and Information Technology, 2020, 20(2): 61-68.

导出引用管理器 EndNote|Ris|BibTeX

链接本文: http://www.tseit.org.cn/CN/abstract/abstract20011.shtml

http://www.tseit.org.cn/CN/Y2020/V20/I2/61

参考文献

[1] 李一凡. 高速铁路周界入侵报警系统关键技术研究及应用[J]. 铁道通信信号, 2017, 53(7): 53-55, 85. [LI Y F. Study and application of key technologies of perimeter intrusion alarming system for high speed railway[J]. Railway Signalling & Communication, 2017, 53(7): 53-55, 85.]

[2] 关晟. 高速铁路周界防护技术研究[J]. 高速铁路技术, 2015, 6(2): 63-67. [GUAN S. Technology research on perimeter protection on high-speed railway[J]. High Speed Railway Technology, 2015, 6(2): 63-67.]

[3] CATALANO A, BRUNO F, PISCO M, et al. An intrusion detection system for the protection of railway assets using fiber bragg grating sensors[J]. Sensors, 2014, 14 (10): 18268-18285.

[4] 牛宏侠, 张肇鑫, 宁正, 等. 铁路轨道异物完整性检测与跟踪算法研究[J]. 交通运输系统工程与信息, 2019, 19(1): 45-54. [NIU H X, ZHANG Z X, NING Z, et al. Detection and tracking algorithm of foreign integrity in railway tracks[J]. Journal of Transportation System Engineering and Information Technology, 2019, 19(1): 45-54.]

[5] 李荣增. 基于视频识别的铁路入侵检测研究与实现 [D]. 武汉: 武汉理工大学, 2016. [LI R Z. Research and implementation of railway intrusion detection based on video recognition[D]. Wuhan: Wuhan University of Technology, 2016]

[6] BILODEAU G A, JODOIN J P, SAUNIER N. 2013 international conference on computer and robot visionchange detection in feature space using local binary similarity patterns[C]// International Conference on Computer & Robot Vision, IEEE Computer Society, 2013: 106-112.

[7] ST-CHARLES P L, BILODEAU G A, BERGEVIN R. A self- adjusting approach to change detection based on background word consensus[C]// 2015 IEEE Winter Conference on Applications of Computer Vision (WACV), IEEE Computer Society, 2015: 990-997.

[8] ST-CHARLES P L, BILODEAU G A, BERGEVIN R. SuBSENSE: A universal change detection method with local adaptive sensitivity[J]. IEEE Transactions on Image Processing, 2014, 24(1): 359-373.

[9] CHEN K, PANG J, WANG J, et al. Hybrid task cascade for instance segmentation[J]. 2019, arXiv: 1901.07518.

[10] LIN T Y, DOLLAR P, GIRSHICK R, et al. Feature pyramid networks for object detection[C]// 2017 IEEE Conference on Computer Vision and Pattern Recognition (CVPR), IEEE Computer Society, 2017: 2117-2125.

[11] CHEN L C, PAPANDREOU G, SCHROFF F, et al. Rethinking atrous convolution for semantic image segmentation[J]. 2017, arXiv:1706.05587.

[12] ANG L L, HACER Y K. Foreground segmentation using convolutional neural networks for multiscale feature encoding[J]. Pattern Recognition Letters, 2018(112): 256-262.

[13] WANG Y, LUO Z, JODOIN P M. Interactive deep learning method for segmenting moving objects[J]. Pattern Recognition Letters, 2016(96): 66-75.

[14] LONG A L, KELES H Y. Learning multi-scale features for foreground segmentation[J]. 2018, arXiv:1808.01477.

[1]	耿庆桥, 贾元华, 陈军团. 基于变权可拓物元的铁路“走出去”公私合营项目风险决策[J]. 交通运输系统工程与信息, 2022, 22(2): 64-71.
[2]	戚建国, 周厚盛, 杨立兴, 周亚茹, 张春田. 灵活编组条件下轨道交通客货协同运输方案优化[J]. 交通运输系统工程与信息, 2022, 22(2): 197-205.
[3]	张晓明, 徐芳, 江文辉, 蒋朝哲. 考虑资金约束的中欧班列运价与运量决策分析[J]. 交通运输系统工程与信息, 2022, 22(1): 30-36.
[4]	张春田, 戚建国, 杨立兴, 高原, 高自友. 基于不确定旅客需求的高速铁路鲁棒列车开行方案研究[J]. 交通运输系统工程与信息, 2022, 22(1): 115-123.
[5]	周勇, 张杰, 钟祾充, 李文锋, 王国栋. 铁路集装箱中心站多轨道吊柔性协同调度优化[J]. 交通运输系统工程与信息, 2022, 22(1): 133-141.
[6]	徐东伟, 彭航, 商学天, 魏臣臣, 杨艳芳. 基于图自编码-生成对抗网络的路网数据修复[J]. 交通运输系统工程与信息, 2021, 21(6): 33-41.
[7]	贺正冰, 徐瑞康, 谢东繁, 宗芳, 钟任新. 数据驱动跟驰模型综述[J]. 交通运输系统工程与信息, 2021, 21(5): 102-113.
[8]	牛惠民. 轨道列车时刻表问题研究综述[J]. 交通运输系统工程与信息, 2021, 21(5): 114-124.
[9]	黄士琛，邵春福，李娟，张小雨，钱剑培. 基于深度学习的车辆轨迹重建与异常轨迹识别[J]. 交通运输系统工程与信息, 2021, 21(3): 47-54.
[10]	李桃迎，王婷，张羽琪. 考虑多特征的高速公路交通流预测模型[J]. 交通运输系统工程与信息, 2021, 21(3): 101-111.
[11]	刘葛辉，陈绍宽，刘爽，金华，王丹阳. 资源约束下城市轨道交通基础设施维修任务安排优化模型[J]. 交通运输系统工程与信息, 2021, 21(3): 163-169.
[12]	徐智勇. 铁路电分相布设及纵断面协同优化研究[J]. 交通运输系统工程与信息, 2021, 21(3): 170-175.
[13]	钟明轩，乐逸祥，周磊山. 高铁站列车和调车作业计划一体化编制方法[J]. 交通运输系统工程与信息, 2021, 21(3): 176-186.
[14]	吕敏，帅斌，周照玉，张士行，左静，李林卿. 考虑边境站延误风险的中欧班列运输路径决策[J]. 交通运输系统工程与信息, 2021, 21(3): 254-259.
[15]	张蒙蒙，许茂增. 考虑政府补贴决策的中欧班列竞合关系研究[J]. 交通运输系统工程与信息, 2021, 21(2): 16-21.

基于视频深度学习的铁路周界入侵检测算法研究

Railway Perimeter Intrusion Detection Algorithms Based on Video Deep Learning

PDF

可视化

摘要/Abstract

引用本文

使用本文

参考文献

相关文章 15

编辑推荐

Metrics