摘要
提出了一种自适应于可调控摄像机的高速公路交通流检测系统。通过改进的背景建模与跟踪方法精确提取运动目标,利用道路标志线为图像与道路建立坐标映射实现车速与流量检测和交通监控。当场景转换时能及时自动更新参数,在实际应用中有很好的灵活性与稳定性。
A highway traffic flow detection system is brought forward which can adapt to Pan Tilt Zoom(PTZ) camera automatically, Moving objects are abstracted accurately by improved background modeling and tracking method. Road marks are used to establish the coordinate mapping between image and road in order to realize vehicle speed estimation, traffic flow statistic and traffic surveillance. Parameters are updated in time. automatically when scene is changed. The system is proved to be intelligent and robust in practiceal applications.
出处
《电视技术》
北大核心
2008年第6期68-70,92,共4页
Video Engineering
基金
国家"863"计划项目(2006AA01Z115)
国家自然科学基金项目(60472002)
关键词
背景建模
目标跟踪
摄像机模型
车速检测
车流监控
background modeling
object tracking
camera model
vehicle velocity estimation
traffic flow surveillance