摘要
在当今节奏高效率的生活工作中,疲劳是引发诸如交通、医疗等事故的重要因素之一。本文基于Perclos等人眼疲劳判定的算法,研究设计了一个嵌入式疲劳预警系统。通过USB摄像头采集视频图像,采用Haar特征的级联分类器从图像中检测出人脸区域和人眼区域,并用帧差分方法跟踪人眼区域,提取其差分统计特征,结合3个准则判断疲劳状态,并进行预警。本系统的设计旨在使打瞌睡人们如驾驶员,能通过系统及时发觉自己的状态,减小由于疲劳产生的事故的可能性。
In today＇s high-paced life and work efficiency, fatigue is one of the important factors that lead to accidents, such as transportation, health care, etc. We can design a fatigue warning system based on the algorithm, such as the human eye fatigue Perclos determination. We make Haar features detect the face region and eye area from the image by USB camera,track the eye area by frame difference, extract its differential statistical features with three criteria determining fatigue. The system is designed to make people, such as drivers, who fall asleep at any time through the system can find their conditions,reducing the possibilities of accidents due to fatigue.
出处
《电子设计工程》
2015年第6期172-175,共4页
Electronic Design Engineering
关键词
预警系统
人眼运动
人眼检测
疲劳
early-warning system
eye movement
eye detection
the fatigue
作者简介
张志文（1957-），男，陕西西安人，教授。研究方向：计算机测控技术、智能化仪表。
