基于目标检测的疲劳驾驶监评系统设计  

Design of Fatigue Driving Supervision and Evaluation System Based on Object Detection

作　　者：代世佳 陈兴文 Dai Shijia;Chen Xingwen(School of Information and Communication Engineering,Dalian Minzu University,Dalian Liaoning 116600,China)

机构地区：[1]大连民族大学信息与通信工程学院,辽宁大连116600

出　　处：《山西电子技术》2025年第2期26-28,共3页Shanxi Electronic Technology

摘　　要：针对营运驾驶员易出现疲劳驾驶而造成危险,利用摄像头提取驾驶员面部特征;通过Percentage of Eyelid Closure Time(Perclos)模型对驾驶员进行疲劳检测;出现疲劳驾驶后,利用云边协同架构将异常信息发送给驾驶员和管理员。系统利用近红外摄像头改善了传统疲劳检测方法在光线复杂的环境下检测效果差的问题,并增加了驾驶员分析评价功能,对驾驶员的异常情况进行评价反馈。In view of the danger caused by fatigue driving of operating drivers,the camera is used to extract the driver’s facial features.The Percentage of Eyelid Closure Time(Perclos)model is used to detect the fatigue of the driver.After fatigue driving,the cloud-edge collaboration architecture is used to send abnormal information to the driver and administrator.The system uses the near-infrared camera to improve the problem of poor detection effect of traditional fatigue detection methods in complex light environments,and adds the driver analysis and evaluation function to evaluate and feedback the driver’s abnormal situation.

关 键 词：疲劳驾驶 目标检测 云边协同 评价反馈 

分 类 号：TP391.41[自动化与计算机技术—计算机应用技术]