基于人体姿态空时特征的驾驶员疲劳检测

doi:10.16097/j.cnki.1009-6744.2022.05.035

交通运输系统工程与信息 ›› 2022, Vol. 22 ›› Issue (5): 337-344.DOI: 10.16097/j.cnki.1009-6744.2022.05.035

• 工程应用与案例分析 • 上一篇

基于人体姿态空时特征的驾驶员疲劳检测

李泰国^*1，张天策¹，李超²，张英志¹，王英¹

1. 兰州交通大学，自动化与电气工程学院，兰州 730070；2. 陕西省康复医院，运动疗法一科，西安710065

收稿日期:2022-06-17 修回日期:2022-08-01 接受日期:2022-08-10 出版日期:2022-10-25 发布日期:2022-10-22
作者简介:李泰国(1985- )，男，甘肃靖远人，讲师。
基金资助:
甘肃省科技计划；兰州交通大学青年科学基金。

Driver Fatigue Detection Based on Spatial-temporal Features and Human Body Pose

LI Tai-guo^*1, ZHANG Tian-ce¹, LI Chao², ZHANG Ying-zhi¹, WANG Ying¹

1. School of Automation and Electrical Engineering, Lanzhou Jiaotong University, Lanzhou 730070, China; 2. Department I of Kinesitherapy, Shaanxi Kangfu Hospital, Xi'an 710065, China

Received:2022-06-17 Revised:2022-08-01 Accepted:2022-08-10 Online:2022-10-25 Published:2022-10-22
Supported by:
The Science and Technology Program of Gansu Province (21JR7RA303，21JR7RA280)；Lanzhou Jiaotong University Youth Science Foundation Project (2020002)。

摘要/Abstract

摘要： 为减少疲劳驾驶给道路交通带来的安全隐患，本文以驾驶员人体姿态为研究对象，分析驾驶动作变化与驾驶员疲劳状态之间的联系，提出基于空时特征与人体姿态的驾驶员疲劳检测模型。首先，以改进的Simple Baselines网络定位驾驶员骨架关键点(空间特征)；其次，分析驾驶过程中人体姿态的变化特点，依据“高内聚、低耦合”的原则将人体关键点模块化，以此为基础设计多个与驾驶员疲劳驾驶相关的特征表示；最后，引入滑动窗口计算各疲劳特征的离散程度，将其作为长短时记忆网络(时间特征)的输入，从而实现对驾驶员疲劳状态的预测。通过13位被试驾驶人的驾驶行为数据实验结果表明：使用本文提出的基于空-时特征和人体姿态的驾驶员疲劳检测模型可达到97.73%的检测精确率，98.95%的召回率以及98.35%的准确率，表明该检测模型具有可行性。

关键词: 智能交通, 疲劳检测, 空-时特征, 长短时记忆网络, 人体姿态

Abstract: To reduce the potential safety hazards caused by fatigued driving, this paper analyzed the relationship between the driver's body pose and the driver's fatigue state and proposed a driver fatigue detection model based on spatial- temporal features and human body pose. The improved simple- baselines network is used to locate the key points (spatial features) of the driver skeleton. By analyzing the changing characteristics of human pose during driving, the key points of the human body are modularized according to the principle of "high cohesion and low coupling", and multiple feature representations related to driver fatigue driving are designed. Finally, a sliding window is introduced to calculate the discrete degree of each fatigue feature, which is used as the input of the Long-short Term Memory Network (temporal features), to realize the prediction of the driver fatigue state. Through the driving behavior data of 13 test drivers, the experimental results show that the driver fatigue detection model based on spatial-temporal features and human posture proposed in this paper can achieve 97.73% precision, 98.95% recall, and 98.35% accuracy, indicating that the detection model is feasible.

Key words: intelligent transportation, fatigue detection, spatial-temporal features, LSTM network, human pose

中图分类号:

U495

李泰国, 张天策, 李超, 张英志, 王英. 基于人体姿态空时特征的驾驶员疲劳检测[J]. 交通运输系统工程与信息, 2022, 22(5): 337-344.

LI Tai-guo, ZHANG Tian-ce, LI Chao, ZHANG Ying-zhi, WANG Ying. Driver Fatigue Detection Based on Spatial-temporal Features and Human Body Pose[J]. Journal of Transportation Systems Engineering and Information Technology, 2022, 22(5): 337-344.

导出引用管理器 EndNote|Ris|BibTeX

链接本文: http://www.tseit.org.cn/CN/10.16097/j.cnki.1009-6744.2022.05.035

http://www.tseit.org.cn/CN/Y2022/V22/I5/337

参考文献

[1] World Health Organization. Global status report on road safety 2015[M]. Geneva: World Health Organization, 2015.

[2] SIKANDER G, ANWAR S. Driver fatigue detection systems: A review[J]. IEEE Transactions on Intelligent Transportation Systems, 2018, 20(6): 2339-2352.

[3] 刘卓凡, 赵霞, 吴付威. 城市道路场景下驾驶注意力需求的影响因素研究[J]. 交通运输系统工程与信息, 2021, 21(1): 132-136. [LIU Z F, ZHAO X, WU F W. Influencing factors of driving attention demand in urban road scenario[J]. Journal of Transportation Systems Engineering and Information Technology, 2021, 21(1): 132-136.]

[4] PERSSON A, JONASSON H, FREDRIKSSON I, et al. Heart rate variability for classification of alert versus sleep deprived drivers in real road driving conditions[J]. IEEE Transactions on Intelligent Transportation Systems, 2020, 22(6): 3316-3325.

[5] HUANG R, WANG Y, LI Z, et al. RF-DCM: MultiGranularity deep convolutional model based on feature recalibration and fusion for driver fatigue detection[J]. IEEE Transactions on Intelligent Transportation Systems, 2020, 23(1): 630-640.

[6] 蔡素贤, 杜超坎, 周思毅, 等. 基于车辆运行数据的疲劳驾驶状态检测[J]. 交通运输系统工程与信息, 2020, 20(4): 77- 82. [CAI S X, DU C K, ZHOU S Y, et al. Fatigue driving state detection based on vehicle running data[J]. Journal of Transportation Systems Engineering and Information Technology, 2020, 20(4): 77-82.]

[7] MAIOR C B S, DAS CHAGAS MOURA M J, SANTANA J M M, et al. Real-time classification for autonomous drowsiness detection using eye aspect ratio[J]. Expert Systems with Applications, 2020, 158(1): 113505- 113517.

[8] CHIOU C Y, WANG W C, LU S C, et al. Driver monitoring using sparse representation with part-based temporal face descriptors[J]. IEEE Transactions on Intelligent Transportation Systems, 2019, 21(1): 346- 361.

[9] 刘明周, 蒋倩男, 扈静. 基于面部几何特征及手部运动特征的驾驶员疲劳检测[J]. 机械工程学报, 2019, 55 (2): 18-26. [LIU M Z, JIANG Q N, HU J. Based on facial geometric features and hand motion characteristics driver fatigue detection[J]. Journal of Mechanical Engineering, 2019, 55(2): 18-26.]

[10] LI R, FU H, LO W L, et al. Skeleton-based action recognition with key-segment descriptor and temporal step matrix model [J]. IEEE Access, 2019, 7(1): 169782- 169795.

[11] 张晓平, 纪佳慧, 王力, 等. 基于视频的人体异常行为识别与检测方法综述[J]. 控制与决策, 2022, 37(1): 14- 27. [ZHANG X P, JI J H, WANG L, et al. Overview of video based human abnormal behavior recognition and detection methods[J]. Control and Decision, 2022, 37(1): 14-27.]

[12] YANG Y, LIU Y, WANG M, et al. Objective evaluation method of steering comfort based on movement quality evaluation of driver steering maneuver[J]. Chinese Journal of Mechanical Engineering, 2014, 27(5): 1027- 1037.

[13] XIAO B, WU H, WEI Y. Simple baselines for human pose estimation and tracking[C]// Proceedings of the European Conference on Computer Vision (ECCV), 2018: 466-481.

[14] SRINIVASU P N, SIVASAI J G, IJAZ M F, et al. Classification of skin disease using deep learning neural networks with MobileNet V2 and LSTM[J]. Sensors, 2021, 21(8): 2852-2879.

[15] 黄士琛, 邵春福, 李娟, 等. 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. [HUANG S C, SHAO C F, LI J, et al. Vehicle trajectory reconstruction and anomaly detection using deep learning[J]. Journal of Transportation Systems Engineering and Information Technology, 2021, 21(3): 47-54.]

[16] SHAHID A, WILKINSON K, MARCU S, et al. Karolinska sleepiness scale (KSS)[M]//STOP, THAT and One Hundred Other Sleep Scales, Springer, New York, NY, 2011: 209-210.

基于人体姿态空时特征的驾驶员疲劳检测

Driver Fatigue Detection Based on Spatial-temporal Features and Human Body Pose

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

[1]	杨柳, 杨莹, 宋允洲, 张宇. 驾驶压力影响因素与识别方法研究综述[J]. 交通运输系统工程与信息, 2022, 22(6): 40-50.
[2]	项新建, 胡海斌, 姚佳娜, 丁祎, 郑永平, 金立. 一种基于物影匹配算法的道路小目标跟踪方法[J]. 交通运输系统工程与信息, 2022, 22(6): 85-94.
[3]	陈志军, 胡军楠, 冷姚, 钱闯, 吴超仲. 基于轻量化网络和注意力机制的智能车快速目标识别方法[J]. 交通运输系统工程与信息, 2022, 22(6): 105-113.
[4]	张文娟, 杨皓哲, 张彬, 李秀杰. 考虑多时间尺度特征的城市轨道交通短时客流量预测模型[J]. 交通运输系统工程与信息, 2022, 22(6): 212-223.
[5]	刘江, 谭思伦, 蔡伯根, 王剑. 基于车路信息交互的车辆卫星定位协同定权方法[J]. 交通运输系统工程与信息, 2022, 22(5): 85-96.
[6]	张名芳, 吴禹峰, 王力, 王庞伟. 基于金字塔特征融合的二阶段三维点云车辆检测[J]. 交通运输系统工程与信息, 2022, 22(5): 107-116.
[7]	张泽锡, 钟文健, 林柏梁. 带时间窗的卡车编队路径优化[J]. 交通运输系统工程与信息, 2022, 22(5): 253-263.
[8]	赵建东, 赵志敏, 屈云超, 谢东繁, 孙会君. 轨迹数据驱动的车辆换道意图识别研究[J]. 交通运输系统工程与信息, 2022, 22(4): 63-71.
[9]	陈喜群, 曹震, 莫栋. 融合卡尔曼滤波的高速公路状态估计误差界限分析[J]. 交通运输系统工程与信息, 2022, 22(4): 72-78.
[10]	张佩瑜, 周建山, 田大新. 网联车辆编队的鲁棒模型预测控制方法[J]. 交通运输系统工程与信息, 2022, 22(4): 268-274.
[11]	秦严严, 朱宜文, 朱立, 唐鸿辉. 智能网联卡车车队混合流通行能力分析方法[J]. 交通运输系统工程与信息, 2022, 22(4): 275-282.
[12]	齐尧, 朱彦齐, 李永乐, 徐友春. 面向动静态混合环境的智能车运动规划方法[J]. 交通运输系统工程与信息, 2022, 22(4): 293-301.
[13]	张仕杰, 唐涛, 刘金涛, 李辰岭. 基于复杂网络的列车辅助驾驶危险致因传播模型[J]. 交通运输系统工程与信息, 2022, 22(4): 129-136.
[14]	李熙莹, 陈丽娟. 一种基于融合网络的慢行交通速度计算方法[J]. 交通运输系统工程与信息, 2022, 22(4): 186-193.
[15]	杨枫, 种大双. 应急车辆动态路径选择的两阶段优化模型[J]. 交通运输系统工程与信息, 2022, 22(3): 84-92.