基于粒子滤波的相关跟踪算法研究  

Research of Correlation Tracking Algorithm Based on Particle Filtering

作　　者：张崇友[1] 董慧颖[1] 兰利宝[1] 

机构地区：[1]沈阳理工大学信息科学与工程学院,辽宁沈阳110168

出　　处：《沈阳理工大学学报》2008年第1期6-9,共4页Journal of Shenyang Ligong University

基　　金：国防预研基金项目

摘　　要：粒子滤波是一种求解后验概率的实用性算法,它通过非参数化的蒙特卡罗模拟方法实现递推贝叶斯滤波.在分析了基于最小平均绝对差值函数的相关跟踪算法的基础上研究了一种基于粒子滤波的相关跟踪算法.并将这两种算法应用于运动目标的跟踪,实现了对道路上运动小汽车的跟踪.最后将两种跟踪算法的仿真实验结果进行了比较,实验结果表明,基于粒子滤波的相关跟踪算法受障碍物遮挡的影响较小,提高了跟踪的鲁棒性.Particle filtering is a practical method for computing posterior PDF （probability density function）, It aims to implement recursive Bayesian filtering through non-parametric Monte Carlo simulation method. The correlation tracking algorithm, based on the function of the minimum average absolute difference is analyzed, and a correlation tracking algorithm based on particle filtering is discussed. Then these two algorithms are used in tracking mov- ing targets, accomplishing track of the car running on the road. Finally, the comparison of two algorithm＇s performance in target tracking is presented. The experimental results show that correlation tracking algorithm based on the particle filtering can improve the robustness of the target tracking.

关 键 词：粒子滤波 相关跟踪 最小平均绝对差值函数 

分 类 号：TP393[自动化与计算机技术—计算机应用技术]