基于云计算的多特征疲劳驾驶检测系统研究与设计

Cloud Computing-based Multi-feature Driver Fatigue Detection System Research and Design

随着私家车的普及,人们对汽车安全性、舒适性要求不断提高,通过对当前车载系统分析和汽车驾驶员疲劳驾驶状态研究,提出了一种基于信息融合的多特征疲劳驾驶检测方案;方案采用高性能嵌入式系统平台与云计算相结合的方式,首先,通过嵌入式系统采集驾驶员面部图像;然后,将数据传输到Face++云计算平台,分析当前驾驶人员身份、年龄与微笑程度;最后,采用数字图像处理技术计算驾驶员头部位移以及统计眼睛眨动规律,综合3种指标预测驾驶员是否处于疲劳状态,实时监测驾驶员驾驶全过程;当检测到驾驶员处于疲劳驾驶状态,则通过语音的方式提醒驾驶员注意行车安全、谨慎驾驶;测试结果表明:该方案检测精度高、实时性强,并且易于和车载系统整合并推广使用。

With the popularity of private cars, people on vehicle safety, comfort requirements continue to increase. This paper presents an approach based on inl＇ormation fusion of multi--feature driver fatigue detection scheme, based on the current automotive systems analysis and research car driver fatigue driving condition. Program uses high--performance embedded system platform with a combination of cloud computing. First of all, the acquisition of the driver＇s facial image through embeded system; and then transfer the data to Faced-d- cloud computing platform, to analyze the current driver status, age and smile level; finally, the use of digital image processing techniques to calcu- late the displacement of the driver＇s head and eye blinking statistical regularity, integrated three kinds of indicators to predict whether the driver is in a state of fatigue, real--time monitoring of the driver to drive the whole process. When the system detects the driver for driving in a state of fatigue, it uses voice prompts alert drivers pay attention to traffic safety and careful driving. The test results show that the scheme of high detection precision, real strong, and easy to integrate with the vehicle system and promotion.