摘要
疲劳驾驶是导致交通事故的重要原因之一。为检测识别驾驶员疲劳状态,根据人的眼动行为存在随机性及模糊性特点,采用不确定性的云模型对眼动特征进行数据处理,构建二维多规则推理生成器检测驾驶员疲劳状况,以此疲劳检测模型为基础构建基于安卓的疲劳预警系统。系统通过手机摄像头实时采集驾驶员面部数据,通过人脸人眼定位后,计算出per-clos和眨眼时间均值。将数据输入疲劳检测模块,一旦检测到驾驶员疲劳,系统即进行文字和语音提醒。该系统成本较低,实时性较好,在模拟驾驶环境下检测率可达到73.98%。
Fatigue driving is one of the causes of traffic fatalities.In order to detect and identify driving fatigue,by considering the ran⁃domness and fuzziness of human eye movement behavior,the uncertain cloud model is used to process eye movement characteristics,and a two-dimensional multi-rule reasoning generator is constructed as driver fatigue detection.Based on this fatigue detection model,an Android-based fatigue early warning system is constructed.The system collects real-time drivers’facial data through mobile phone camera,and calculates per-clos and blink time mean after facial eye location.Their data are input into the fatigue detection module.Once driver fatigue is detected,the system can carry out text and voice reminders.The system has low cost and good real-time perfor⁃mance.
作者
陈瑜
李锦涛
徐军莉
陈威月
CHEN Yu;LI Jin-tao;XU Jun-li;CHEN Wei-yue(School of Information Engineering,Jiangxi University of Technology;Collaborative Innovation Center,Jiangxi University of Technology,Nanchang 330098,China)
出处
《软件导刊》
2020年第5期116-119,共4页
Software Guide
基金
国家级大学生创新创业训练计划项目(201810846002)
江西省教育厅科技项目(GJJ180979)。
关键词
眼动特征
疲劳驾驶
预警系统
定性推理器
云模型
eye movement characteristics
fatigue driving
early warning system
qualitative reasoner
cloud model
