基于模糊控制的驾驶疲劳检测

提出了一种判断疲劳程度的新方法,通过检测眼睛、嘴巴状态和头部位置等能够反应疲劳的生理特征,利用模糊控制器推理,得到人的疲劳状态的数值表示,改善了疲劳或者非疲劳的二值表示形式.通过肤色识别和阚值选取等方法得到人眼检测,进行边界提取的结果优于边缘检测.利用fisher线性分类器进行嘴唇和肤色的分类,提高了检测速度.模糊推理更精确的反应了人的疲劳程度,实验结果表明了检测方法的有效性和可信性.

Application of fuzzy-control in detection of fatigued drivers

A new method of fatigue degree evaluation is presented. Fatigue characteristics are obtained by detecting the eyes, mouth and head, and utilize fuzzy control system to train. It is better than the ways that uses fatigue or not fatigue as results. Skin identify combined with threshold are used to detect the eye and the result is better than edge detection function. Using fisher classifier to detect mouth have saved time. Experiments shows that detect ways turn out to be effectively and credibility.