基于GM-HMM的驾驶人疲劳状态检测

删除或更新信息，请邮件至freekaoyan#163.com(#换成@)

本站小编 Free考研考试/2024-01-16

-->

张明恒,刘朝阳,郭政先,万星.基于GM-HMM的驾驶人疲劳状态检测[J].,2021,61(4):

基于GM-HMM的驾驶人疲劳状态检测

Driver fatigue state detection based on GM-HMM

DOI：10.7511/dllgxb202104009

中文关键词:舱内感知技术(ICS)驾驶疲劳 GM-HMM 脑电信号

英文关键词:in-cabin sensing (ICS)driving fatigue GM-HMM electroencephalogram signal

基金项目:国家自然科学基金资助项目(51675077)；中国博士后科学基金资助项目(2015M5813292017T100178).

作者	单位
张明恒,刘朝阳,郭政先,万星

摘要点击次数:201

全文下载次数:225

中文摘要:

驾驶人状态检测是舱内感知技术(ICS)的重点研究方向之一，其中驾驶疲劳作为交通事故致因的重要组成部分，越来越受到人们的重点关注．驾驶人疲劳检测的本质是通过相关特征对当前驾驶人状态的间接评估过程，其中疲劳状态的标定对构建特征-疲劳状态的映射关系具有重要影响，也是目前相关车载系统研发所面临的共性关键问题．由此，基于脑电(EEG)信号数据和驾驶疲劳的动态生成特性提出了一种高斯混合隐马尔可夫模型(GM-HMM)进行疲劳状态评估，以对相关车载系统研发提供必要的疲劳状态比对参考．实验和对比测试结果表明，所提模型在准确率、灵敏度和特异性方面具有较大优势．

英文摘要:

Driver state detection is one of the key areas of development for in cabin sensing (ICS). And as an important part of the causes of accidents, driving fatigue has increasingly received focused attention. Essentially, driver fatigue detection is an indirect process of assessing the driver state through relevant features. Among them, fatigue state calibration is significant for establishing the relationships between fatigue and features, and it is also a common key problem faced by the current research and development of related on board systems. Therefore, based on electroencephalogram (EEG) signal data and the dynamic generation characteristics of driving fatigue, a Gaussian mixture hidden Markov model (GM-HMM) is proposed to assess fatigue state, which can provide necessary references for the driver fatigue detection research of on board systems. The experimental and comparative results show that the proposed model has advantages over other related models in terms of accuracy, sensitivity and specificity.

查看全文查看/发表评论下载PDF阅读器

关闭

基于GM-HMM的驾驶人疲劳状态检测

本站小编 Free考研考试/2024-01-16

相关话题/

领限时大额优惠券,享本站正版考研考试资料!