基于Odroid的疲劳驾驶检测研究

Research on Fatigue Driving Detection Based on Odroid System

基于眼睛闭合状态识别的方法设计了一种嵌入式疲劳检测平台,能够实时采集驾驶员图像并进行疲劳检测。方法:疲劳检测的步骤为:首先用摄像头实时采集驾驶员图像,并对获取的图像进行人脸检测;然后,在识别出的人脸区域进行人眼位置检测,进而识别眼睛闭合状态;最后根据Perclos疲劳判定标准判别驾驶员是否疲劳,如果发现疲劳给予报警。最终将计算机上开发的疲劳检测程序移植到嵌入式设备,通过Qt开发用户界面,实现了嵌入式平台的疲劳检测。

This paper designs an embedded fatigue detection device based on the eye closed state recognition method which can collect the driver image and fatigue detection in real time.Methods:The steps of fatigue detection are as follow:First,the image of driver is collected using real-time camera;Second,the face is found from the image by face detection method,and the eye position is located from the face area the eye state is recognized;Finally,after detecting the open or closing state of the eyes,the Perclos fatigue criterion is used to determine whether the driver is fatigue driving,and the alarm is given if fatigue is found.This design transplants the fatigue detection procedures from computer to embedded device.Though develop the UI by Qt,the whole embedded fatigue detection system is realized.