摘要
研究了基于视频的交通事件自动检测技术,它通过车辆跟踪和分析车辆的运动特征来判断所发生的交通事件。针对三种常见的交通事件车辆逆行、车辆停止和堵车进行建模。实验证明,该检测方法具有很好的实用效果。
In this paper, video--based automatic incident detection technology is introduced, which is through the vehicle tracking and the motion features analysis of vehicles to determine the traffic incidents. According to three common traffic incidents, such as vehicle retrograde, vehicle stop and traffic jam, the modeling is made. The test results indicate that the detection method has good practical effect.
出处
《电视技术》
北大核心
2010年第11期101-103,116,共4页
Video Engineering
基金
国家自然科学基金项目(6087207360975017)
江苏省自然科学基金项目(BK2008291)
关键词
事件检测
逆行
停车
堵车
incident detection
vehicle retrograde
vehicle stopped
traffic jam
