疲劳状态检测系统的设计与实现

Design and Implementation of Fatigue State Detection System

为了降低因疲劳作业诱发的各类安全风险,开发一套高性能的疲劳检测预警系统是十分必要的。本文设计并开发了一套基于面部识别的疲劳状态检测系统,利用Dlib官方预训练的人脸特征位置检测器获取被监测人员面部68个特征点及坐标,通过坐标计算得到的眼睛、嘴巴纵横比以捕获眨眼、打哈欠动作;同时,通过头部姿态估计算法捕获瞌睡点头动作。最终系统根据眨眼、哈欠及点头的频率综合判断被监测人员是否处于疲劳状态并作出安全提示。实验的结果表明,该疲劳状态检测系统能实时、准确地检测被监测人员的疲劳状态,并及时发出警报。

In order to reduce all kinds of safety risks caused by fatigue operation, it is necessary to develop a set of high performance fatigue detection and early warning system. In this paper, a fatigue state detection system based on facial recognition was designed and developed. 68 feature points and coordinates were obtained by using Dlib official pre-trained face feature position detector, and the aspect ratio of eyes and mouth was calculated to capture blinking and yawning movements. Meanwhile, nodding motion was captured by head pose estimation algorithm. Finally, the system will comprehensively judge whether the monitored person is in a state of fatigue according to the frequency of blinking, yawning and nodding and give safety tips. The experimental results show that the fatigue state detection system can detect the fatigue state of the monitored personnel in real time and accurately, and give an alarm in time.