【作者】 王景中； 刘凯；

【Author】 Jing Zhong WANG~1,Liu Kai~2 1.North China University of Technology,Beijing,China 2.North China University of Technology,Beijing,China

【机构】 北方工业大学；

【摘要】 针对传统Meanshift算法跟踪运动目标容易丢失目标的情况,利用一种Kalman滤波和窗口自适应的Meanshift算法以及三步搜索法相结合的改进算法。Kalman滤波可以克服复杂背景的干扰,三步搜索法能够方便的在目标被遮挡时快速重获目标,窗口自适应的Meanshift算法可以更好的避免背景的干扰,通过改变窗口尺寸介绍背景的介入。本文利用该方法实习成像系统中目标跟踪模块。更多还原

【Abstract】 Traditional Meanshift algorithm for tracking moving targets is easy to fail,so the use of a Kalman filter and window adaptive Meanshift algorithm by means of the combination of three-stage search algorithm are proposed.Kalman filter can overcome complex background interference,three-step search method can easily and quickly regain the moving targets, and window adaptive Meanshift algorithm can better avoid background interference,introduced by changing the background window size of intervention.In this paper,we use this method in an imaging system in which there is a tracking module.更多还原