驾驶员疲劳驾驶预警系统的设计

马乐^1,2, 姜立标³, 王会荣⁴, 王蒙⁵

1.中国农业大学工学院;2.华南理工大学广州汽车学院;3.华南理工大学机械与汽车工程学院;4.哈尔滨工业大学(威海)汽车工程学院;5.Griffith College Dublin



摘要:

针对疲劳驾驶监测的要求,设计了一种基于TI公司高速图像处理芯片TMS320DM642的疲劳驾驶预警系统,用以判定驾驶员的疲劳程度,降低交通事故.系统采用机器视觉的方法,先通过CCD摄像头拍下驾驶员的面部图像序列,然后检测出脸部,从而定位眼睛,再利用PERCLOS算法判定疲劳状态.实验结果表明,该系统准确率高、速度快,可以满足非接触式、全天候、实时监测的要求.

关键词:  疲劳驾驶  数字信号处理器  人脸检测  PERCLOS算法  级联分类器

DOI：10.11918/j.issn.0367-6234.2011.05.029

分类号:TP391.41

基金项目:广东省自然科学基金资助项目(9451064101003049)



Design of driver drowsiness detection system

MA Le^1,2, JIANG Li-biao³, WANG Hui-rong⁴, WANG Meng⁵

1.College of Engineering of China Agriculture University,100083 Beijing,China;2.Guangzhou Auto College,South China University of Technology,510800 Guangzhou,China;3.School of Mechanical & Automotive Engineering,South China University of Technology,510640 Guangzhou,China;4.School of Automotive Engineering,Harbin Institute of Technology At Weihai,264209 Weihai Shandong,China;5.Griffith College Dublin,Dublin,Ireland

Abstract:

Aiming at the demands of fatigue driving monitoring,a drowsiness detection system based on High-speed DSP TMS320DM642 from TI Inc was designed.It can determine the driver’s level of attention,and reduce the number of traffic accident.This system uses the machine vision method to get video sequences of a driver from the CCD camera,and than detects the face,eyelid to determine the state of drowsiness by the PERCLOS criterion.The experimental results show that this system has higher accuracy and speed,and can satisfy the demands of non-contact,all illumination condition and real-time monitoring.

Key words:  fatigue driving  DSP  face detection  PERCLOS criterion  cascade