改进粒子滤波算法在目标跟踪中的应用实现  

The Application and Realization of Target Tracking Based on Improved Particle Filter Algorithm

作　　者：曲浩添 苏霄[2] 田景文[2] 

机构地区：[1]中石油煤层气有限责任公司,北京100028 [2]北京联合大学信息学院,北京100101

出　　处：《自动化技术与应用》2011年第8期43-47,共5页Techniques of Automation and Applications

摘　　要：针对移动目标跟踪的非线性、非高斯的特点,本文系统介绍了基于ARM11的嵌入式设备进行移动目标跟踪的应用实现。核心应用算法使用改进的粒子滤波算法,其中粒子滤波算法的改进采用对粒子加权以及重新采样,以克服样本贫化现象和区分粒子的重要性程度。然后阐述了将粒子滤波算法移植到嵌入式设备以实现移动目标跟踪的应用需要。Aiming at the characteristics of nonlinear, non-gaussian distribution of the moving target tracking, this paper introducs an applied realization about move target tracking base on embedded device of ARM11. The core applied algorithm comes from the improved particle filter algorithm. Though adopting particles weighted and resampling. It can overcome the sample impoverishment phenomenon and distinguish the importance of particles. This article describes the needs of using this particle filter algorithm to track the moving target on embedded equipment.

关 键 词：粒子滤波算法 目标跟踪 嵌入式系统移植 

分 类 号：TP18[自动化与计算机技术—控制理论与控制工程]