基于眼睛状态的驾驶员疲劳模糊识别研究  

A Fuzzy Recognition Algorithm for Driver's Fatigue Based on Eyes State

作　　者：翁茂荣[1] 

机构地区：[1]浙江工贸职业技术学院汽车与机电工程系,浙江温州325003

出　　处：《计算机仿真》2008年第6期244-247,273,共5页Computer Simulation

基　　金：温州市2005年第一期科技计划项目(G2005036)

摘　　要：首先介绍了当前国内外各种驾驶员疲劳识别方法,并分析其优缺点后,通过对驾驶员眼睛状态的定位和跟踪,提取眼睑距离来得到眨眼的频率和间隔,提出了驾驶员疲劳的模糊识别算法判定驾驶员的疲劳程度,并利用摄像机实时拍摄驾驶员的脸部,试验表明,提出的模糊识别方法能快速、准确地判断甚至预测瞌睡的发生,对于一般驾驶时目标检测正确率可以达到95%,当驾驶员头部开始转动时,正确率会下降至87%,但仍在可以判别范围内,具有较快的响应速度和鲁棒性。The current methods and technologies for driver fatigue detection are investigated and contrasted in this paper, and the merits and shortcomings of these techniques are analyzed. Then blink frequent and alternation are gained by the way of distance of eyelid based on eye orientation and track. A fuzzy recognition algorithm for driver＇s fatigue is put forward, which is cue for determining driver＇s fatigue state. A camera was laid on the pane/to shoot the driver＇s face. Experiment results showed the this method can quickly and effectively identify driver＇s fatigue, the target detection recognition is 95%, and if the driver＇s head sways, the target detection recognition is 87% ,which still has preferable response speed and robustness.

关 键 词：驾驶员疲劳 眼部定位 特征提取 模糊识别 

分 类 号：U461.91[机械工程—车辆工程]