基于mean-shift算法的目标跟踪方法

针对雷达多目标跟踪提出一种基于mean-shift算法的目标跟踪方法.首次将mean-shift的方法应用于目标的数据关联,找出源于目标的观测值后对其进行Kalman滤波,从而估计出目标运动的轨迹,实现目标跟踪.MST跳出传统思维框架,首次利用概率密度分布的不同来区分服从不同参数分布的数据,从整体上对观测数据进行整合再结合最邻近法完成数据关联,该方法具有计算速度快,跟踪效果好的特点.

A Target Tracking Method Based on Mean Shift Algrothm

A target tracker based on the mean-shift algorithm is proposed.The MST(Mean-Shift Tracker)employed the mean-shift algorithm is applied to target data associations and then Kalman filter is used to estimate the track of the targets.In order to tack targets,the MST utilizes the data of different probability density functions to distinguish them.This idea is very different from the traditional methods.By doing so,the data from the targets and clutter background having different probability density functions can be differentiated,whereas the data associations of the target tracks are carried out by combining the target probability density functions and the near neighbor data association.The Simulation results show that the computing load and the response time delay of the algorithm is significantly low.Furthermore,a good track performance is achieved.