摘要
建立了基于摄像机、图像采集卡及计算机的车辆跟踪硬件系统。研究了基于运动矢量的车辆跟踪算法。作者首先进行背景初始化,然后,利用背景差法和检测线法对通过目标交通场景的车辆进行检测,进而根据三步法计算车辆的运动矢量,并进行车辆位置的更新和遮挡处理,实现了车辆的实时跟踪。
A vehicle tracking hardware system consists of video camera, image acquisition card and microcomputer was constructed. Meanwhile, a vehicle tracking algorithm based on motion vector was studied. We firstly initialized the background by calculating the acquired images, and used background difference and detection line methods to detect the vehicles passing through the object traffic scene; then, calculated the motion vectors using a three step method and updated the vehicles＇ location continuously; at last, the occlusion is processed also. The result showed that the goal of real-time vehicle tracking was achieved.
出处
《交通运输工程与信息学报》
2007年第4期110-115,共6页
Journal of Transportation Engineering and Information
基金
高等学校科技创新工程重大项目培育资金项目(705020)
江苏省自然科学基金项目(BK2004077)
关键词
车辆跟踪
交通信息采集
视频检测
Vehicle tracking, traffic information collection, video detection
