一种基于卡尔曼滤波的运动物体跟踪算法

针对实时视频监控领域中传统的Camshift算法不能有效解决遮挡和高速运动等问题，提出一种改进的Camshift算法与卡尔曼滤波相结合的运动物体跟踪算法。首先，通过二次搜索来调整搜索窗口的位置和大小，保证Camshift跟踪的可靠性；然后，在Camshift算法的基础上通过卡尔曼滤波对搜索窗口进行运动预测，保证实时跟踪。实验表明该方法具有较好的实时性，并能够有效地解决遮挡等问题。

Algorithm for moving object tracking based on Kalman filter

Based on the situation that the classic Camshift algorithm could not effective solve the problems such as target was covered and high speed moving in the field of real-time video tracking, this paper proposed a improved moving object tracking algorithm that combinating Camshift algorithm and Kalman filter. Firstly, in order to improve tracking stability of Camshift, adjusted the places and sizes by twice searching and then, in order to ensure real-time tracking, implemented motion prediction in search window by Kalman filter on the base of Camshift algorithms. The experimental results show that the improved algorithm has better real-time and can solve the cover problem more effective.