基于机器视觉的行车安全综合保障系统研究  

Comprehensive Safety Guarantee System of Moving Vehicle Based on Machine Vision

作　　者：王荣本[1] 郭烈[1] 顾柏园[1] 余天洪[1] 张明恒[1] 

机构地区：[1]吉林大学交通学院,吉林长春130025

出　　处：《山东交通学院学报》2006年第2期1-6,共6页Journal of Shandong Jiaotong University

摘　　要：驾驶员疲劳驾驶或精神分散容易导致发生车道偏离以及追尾碰撞等恶性交通事故,利用机器视觉技术监测驾驶员的眼部、嘴部、头部等器官的运动特征,建立基于驾驶员面部视觉信息的疲劳与精神分散状态综合量化评价模型,实现实时有效的动态监测和预警,并系统研究了本车前方车辆探测方法和安全车距的评定准则,实现防追尾碰撞预警。初步试验表明:当驾驶员疲劳驾驶、无意识车道偏离以及行车安全车距过小时,该行车安全综合保障系统能有效地对驾驶员进行警示。此外,根据行车安全保障技术领域的研究现状,还开展了道路上行人检测的技术研究,并取得初步成果。In order to effectively reduce the severe truffle accidents such as lane departure and collisions, which are caused by driver＇s fatigue and spirit distraction, this paper proposes a method based on machine vision to monitor the moving characteristics of driver＇s eyes, mouth and head, and establishes a comprehensive evaluation model based on driver＇s face vision information and spirit distraction. It can realize real - time monitoring and pre - warning effectively and dynamically. Besides, the paper studies on the estimation criterion of the front vehicles and safety distances, so as to realize the pre - warning of collisions. Experiments show that when the drivers drive with fatigue, deviate from the lane unconsciously or when the safety distance is too short, the comprehensive safety guarantee system can realize pre - warning effectively. Moreover, based on the research situations overseas, the paper also does some research on passersby monitoring technology, and achieves some results.

关 键 词：行车安全 机器视觉 综合保障系统 

分 类 号：U471.15[机械工程—车辆工程]