基于改进PERCLOS的疲劳驾驶检测系统的设计  

Design of fatigue driving detection system based on improved PERCLOS

作　　者：许龙铭 XU Longming(School of Communication Engineering,Guangzhou City University of Technology,Guangzhou 510800,China)

机构地区：[1]广州城市理工学院通信工程学院,广东广州510800

出　　处：《现代电子技术》2023年第22期41-45,共5页Modern Electronics Technique

基　　金：广东省普通高校重点领域专项(2022ZDZX1041)。

摘　　要：针对疲劳驾驶事件频发、检测难度较高的问题,设计一种基于改进PERCLOS的疲劳驾驶检测系统。以搭载CMOS摄像头的OpenMV作为主控,实时检测驾驶人的人脸图像,采用基于Haar特征的Cascade分类器来分割人眼区域,再从人眼区域内提取瞳孔的颜色深度特征,计算得出人眼开合度,从而判断驾驶员是否在眨眼。使用眨眼频率替代PERCLOS算法中的眼睛闭合时间的评价指标,结合方向盘握力特征和汽车横向加速度特征进行多传感器信息融合,综合判断驾驶员是否处于疲劳状态。实验结果表明,系统检测准确率达到90%以上,实时性较强,能够以一种低成本方案解决疲劳驾驶检测的需求。In allusion to the problems of frequent fatigue driving events and high detection difficulty,an improved PERCLOS based fatigue driving detection system is designed.With OpenMV equipped with CMOS camera as the main control,the driver's face image is detected in real time.The Cascade classifier based on the Haar feature is used to segment the eye area,and then the pupil color depth feature is extracted from the eye area to calculate the eye opening and closing degree,so as to determine whether the driver is blinking.The blink frequency is used to replace the evaluation index of eye closing time in PERCLOS algorithm,and the multi-sensor information fusion is conducted by combing the steering wheel grip characteristics and vehicle lateral acceleration characteristics,so as to comprehensively determine whether the driver is in a state of fatigue.The experimental results show that the detection accuracy of the system can reach more than 90%,and the real-time performance is strong,which can solve the demand of fatigue driving detection with a low-cost solution.

关 键 词：PERCLOS 疲劳驾驶 CMOS摄像头 OpenMV Cascade分类器 颜色特征提取 多传感器信息融合 

分 类 号：TN911.23-34[电子电信—通信与信息系统] TP391.4[电子电信—信息与通信工程]