摘要
基于车载疲劳驾驶检测系统的需要,使用手机的前置摄像头设计了用于手机的疲劳驾驶检测系统。通过使用Haar-like特征和Adaboost分类器进行人脸的定位,并在此之上以同样的方法进行人眼的粗略检测。然后通过灰度投影结合Otsu法二值化得到眼睛的精确开度,判断人眼的状态。以此根据PERCLOS方法对驾驶员的疲劳程度进行检测。实验结果表明该系统具有较高的准确度,并对光照的变化有一定的适应性,能够有效地应用到疲劳检测中。
In order to meet the needs of vehicle driver drowsiness detection system,A driving drowsiness detection system for mobile phone is designed by using the front camera of the phone. The Haar-like feature and Adaboost classifier are used to detect the driver's face,and then,same method is used for a rough eye detection. Next,Otsu binarization is combined with gray projection to calculate the accurate eye height and evaluate the statues of eyes. Finally,a method according to the PERCLOS is used to evaluate the degree of drowsiness. Experimental results show that the system has a high accuracy,and it can adapt to the changes of light. So it can be effectively applied to the driver's drowsiness detection.
出处
《信号处理》
CSCD
北大核心
2015年第9期1138-1144,共7页
Journal of Signal Processing
基金
国家自然科学基金资助项目(61171151)
国家重点基础研究发展计划("973"计划)基金资助项目(2012CB316400)