基于脑电信号的驾驶疲劳检测综述  

Review on driving fatigue detection based on EEG

作　　者：王洪涛 殷浩钧 陈创泉 Anastasios Bezerianos WANG Hongtao;YIN Haojun;CHEN Chuangquan;ANASTASIOS Bezerianos(Intelligent Manufacturing Department,Wuyi University,Jiangmen 529020,Guangdong China;N1 Institute for Health,National University of Singapore,Singapore 117456,Singapore;Medical School,University of Patras,26500 Patras,Greece)

机构地区：[1]五邑大学智能制造学部,广东江门529020 [2]新加坡国立大学N1健康研究所,新加坡117456 [3]佩特雷大学医学院,希腊佩特雷26500

出　　处：《华中科技大学学报（自然科学版）》2022年第11期54-65,78,共13页Journal of Huazhong University of Science and Technology(Natural Science Edition)

基　　金：广东省教育厅重点领域专项资助项目(2020ZDZX3018);广东省教育厅特色创新类资助项目(2021KTSCX136);广东省科技厅科技发展专项资助项目(2020182);五邑大学港澳联合研发资助项目(2019WGALH16);广东省基础与应用基础研究基金资助项目(2020A1515111154);广东省生物医学信息检测与超声成像重点实验室开放课题(SZD201909)。

摘　　要：围绕基于脑电信号的驾驶疲劳检测,通过大量文献检索,总结了脑电信号采集设备、脑电信号特征提取方法和脑电信号分类方法三个方面现状.分析了采集设备的便携性与舒适度问题、与疲劳相关特征的稳定性问题及疲劳检测模型的鲁棒性问题,进而梳理并总结出基于脑电信号驾驶疲劳检测的三个发展趋势:从湿式电极到干式电极;从通道内特征到通道间特征;从浅层机器学习到深度学习.Focusing on detecting driving fatigue based on electroencephalogram(EEG) signals,through a significant number of literature searches,the current status of EEG signal acquisition equipment,EEG signal feature extraction methods and EEG signal classification methods were summarized. The issues of the portability and comfort of the acquisition equipment,the stability of fatigue-related features,and the robustness of the fatigue detection models were analyzed. Furthermore,the development trend of EEG-based driving fatigue detection were summarized in three aspects:from wet electrodes to dry electrodes,from intra-channel features to inter-channel features,and from shallow machine learning methods to deep learning methods.

关 键 词：脑电信号 驾驶疲劳 脑功能网络 机器学习 深度学习 

分 类 号：R318[医药卫生—生物医学工程]