卡尔曼滤波器是一个对动态系统的状态序列进行线性最小误差估计的算法，一般用于线性系统。一般在运动跟踪领域中摄像机相对于目标物体运动有时属于非线性系统，但由于在一般运动跟踪问题中图像采集时间间隔较短，可近似将单位时间内目标在图像中的运动看作匀速运动，采用卡尔曼滤波器可以实现对目标运动参数的估计。-Kalman filter is a state sequence of linear dynamic systems smallest error estimation algorithm for linear systems in general. Usually in the field of motion tracking camera motion with respect to the target object may belong to nonlinear systems, but in general due to motion tracking problems shorter image acquisition time interval, a unit of time can be approximated to the target motion in the image is regarded as uniform motion, Kalman filter can achieve the target motion parameter estimation.