一种基于人脸特征点的飞行签派员疲劳检测系统
A Fatigue Detection System for Flight Dispatcher Based on Face Feature Points
高强 1潘俊 2刘薇1
作者信息
- 1. 广州铁路职业技术学院信息工程学院,广东广州 511300
- 2. 广州民航职业技术学民航电子信息工程学院,广东广州 510403
- 折叠
摘要
民航飞行签派员长时间处于工作强度大、注意力高度集中的工作状态,生理、心理都容易产生疲劳,容易导致错误签派而造成飞行安全事故.本文通过摄像头采集签派员工作图像,对采集的图像进行人脸区域及关键点检测,然后对签派员的睁闭眼状态、嘴巴张合度等进行计算得到眼睑纵横比EAR值、嘴部纵横比MAR值,接下来基于EAR、MAR获取PERCLOS、平均闭眼时长和打哈欠频率三项特征并将其进行融合,最后,根据疲劳等级判断标准,利用OpenCV、Dlib等工具开发了签派员疲劳检测系统,该系统能够检测不同程度疲劳状态.
Abstract
Civil aviation flight dispatchers are in a working state with high work intensity and high concentration for a long time.They are prone to physical and psychological fatigue,which can easily lead to wrong dispatches and flight safety accidents.In this paper,the dispatcher's working image is collected by the camera,the face area and key points are detected on the collected image,and then the dispatcher's eye opening and closing state,mouth opening and closing degree,etc.are calculated to obtain the eyelid aspect ratio EAR valueand the mouth aspect ratio MAR value,and then based on EAR and MAR,three features of PERCLOS,average eye closure time,and yawning frequency are obtained and fused,finally,according to the fatigue level judgment standard,the dispatcher fatigue detection system is developed by using OpenCV,Dlib and other tools,which can detect different levels of fatigue.
关键词
人脸特征/签派员/纵横比/PERCLOS/疲劳检测系统Key words
facial features/dispatcher/aspect ratio/PERCLOS/fatigue detection system引用本文复制引用
基金项目
广东省普通高校特色创新项目(2022KTSCX227)
广东省高等教育学会课题(22GYB013)
出版年
2024
NETL
NSTL
万方数据