摘要
鉴于传统的疲劳检测模型通过驾驶员单一疲劳特征检测具有局限性的问题,提出了一种新的驾驶员疲劳检测模型。首先使用改进的AdaBoost算法进行人脸检测,解决复杂光源和背景的影响,提高人脸检测效率。然后用LBF算法进行人眼检测,用三庭五眼法进行嘴部检测,通过人眼高宽比和像素比检测人眼闭合程度,通过嘴部高宽比和圆形度检测嘴部打哈欠状态,再综合眼部疲劳特征计算闭眼时间,利用打哈欠频率计算嘴部疲劳。最后综合上述疲劳特征检测驾驶员疲劳状态。实验表明该方法可有效检测驾驶员疲劳状态,满足疲劳检测系统对实时性、鲁棒性的要求。
In view of the limitation of the traditional fatigue detection model,a new driver fatigue detection model is proposed.Firstly,the improved AdaBoost algorithm is used for face detection to solve the influence of complex light source and background and improve the efficiency of face detection.Then LBF algorithm is used to detect human eyes,and mouth is detected by three-court five-eye method.The degree of human eyes closure is detected by eye height-width ratio and pixel ratio,and the state of mouth yawning is detected by mouth height-width ratio and circular degree,then,the time of closing eyes is calculated by com⁃bining the features of eye fatigue,and the frequency of yawning is used to calculate mouth fatigue.Finally,the driver's fatigue state is detected by synthesizing the above fatigue characteristics.The experimental results show that the method can detect the fatigue state of the driver effectively and meet the requirements of real-time and robustness of the fatigue detection system.
作者
陈立潮
王冠男
CHEN Lichao;WANG Guannan(College of Computer Science and Technology,Taiyuan University of Science and Technology,Taiyuan 030024)
出处
《计算机与数字工程》
2023年第3期721-726,共6页
Computer & Digital Engineering
基金
山西省应用基础研究项目(编号:201801D221179)
太原科技大学校博士科研启动基金项目(编号:20162036)资助。