摘要
随着国家经济的迅速发展,基础设施日益完善,人民生活水平不断提高,安全出行成为一项重大的民生课题。很多交通事故都是由疲劳驾驶引起。目前,快速、精准检测疲劳驾驶成为一个热点研究问题。基于此,从传统疲劳驾驶检测方案的局限性出发,引出了基于深度学习和微表情检测的防疲劳驾驶检测方案,归纳总结了微表情检测流程和深度学习流程。
With the rapid development of national economy, the improvement of infrastructure and the improvement of people's living standards, safe travel has become a major livelihood issue. Many traffic accidents are caused by fatigue driving. At present, rapid and accurate detection of fatigue driving has become a hot research topic. Based on this, starting from the limitations of the traditional fatigue driving detection scheme, the anti-fatigue driving detection scheme based on deep learning and micro-expression detection is introduced, and the micro-expression detection process and deep learning process are summarized.
引文
[1]Haraldsson P O,Akerstedt T.[Drowsiness--greater Traffic Hazard than Alcohol.Causes,Risks and Treatment][J].Lakartidning en,2001,98(25):3018.
[2]Royal D.NATIONAL SURVEY OF SPEEDING,UNSAFE,DISTRACTED AND DROWSY DRIVING ATTITUDESAND BEHAVIOR,2002.VOLUMEIII,METHODS[J].Nutrition Research,2003,23(11):1477-1487.
[3]赵磊.基于深度学习和面部多源动态行为融合的驾驶员疲劳检测方法研究[D].济南:山东大学,2018:154.
[4]Gepperth A R T.Object Detection and Feature Base Learning with Sparse Convolutional Neural Networks[M].Berlin:Springer,2006:221-232.
[5]曾鸽.基于深度学习的驾驶员疲劳状态检测研究[D].长沙:湖南大学,2018:69.