车辆跟踪是智能交通系统中的一项关键技术,在研究现有车辆跟踪算法基础上,提出了一种基于UKF与SIFT的车辆跟踪算法:通过将车辆的外接矩形信息转化为UKF滤波参数,对车辆运动进行建模,结合SIFT特征匹配能够有效地解决车辆遮挡问题;实验结果表明,该方法能够对运动车辆实现稳定的跟踪。并且能够有效地解决车辆遮挡问题。
Vehicle tracking is one of the key technologies in Intelligent Transportation System. In this paper, a new vehicle tracking algorithm based on UKF and SIFT is proposed after researching of the the existing algorithm: vehicle moving model is made by converting the parameters of the contour--rectangles to the parameters of UKF filter, objects sheltering can be effectively resolved through matching SIFT feature. The experiments demonstrate that vehicles can be tracked stably, and objects sheltering can be effectively resolved.