作者投稿和查稿 主编审稿 专家审稿 编委审稿 远程编辑

计算机工程 ›› 2010, Vol. 36 ›› Issue (20): 179-181. doi: 10.3969/j.issn.1000-3428.2010.20.063

• 人工智能及识别技术 • 上一篇    下一篇

基于DSmT与粒子滤波的多传感器融合

夏建明a,b,杨俊安a,b,张 琼a,b   

  1. (解放军电子工程学院 a. 信息系;b. 安徽省电子制约技术重点实验室,合肥 230037)
  • 出版日期:2010-10-20 发布日期:2010-10-18
  • 作者简介:夏建明(1982-),男,硕士研究生,主研方向:信号分析,机器学习;杨俊安,教授、博士、博士生导师;张 琼,硕士研究生
  • 基金资助:

    国家自然科学基金资助项目(60872113)

Multi-sensor Fusion Based on DSmT and Particle Filtering

XIA Jian-minga,b, YANG Jun-ana,b, ZHANG Qionga,b   

  1. (a. Department of Information; b. Key Laboratory of Electronic Restriction Technology of Anhui Province,PLA Electronic Engineering Institute, Hefei 230037, China)
  • Online:2010-10-20 Published:2010-10-18

摘要:

为实现多传感器对机动目标状态的跟踪,提出一种基于DSmT与粒子滤波的多传感器融合算法。在各传感器利用粒子滤波方法处理观测数据的基础上,运用DSmT作为融合工具,将观测数据转化为辨识框架内的元素及其mass值,得到最终融合结果。实验结果表明,该方法可减小距离误差,提高跟踪精度,且运算复杂度能满足在线实时融合的要求。

关键词: DSmT技术, 粒子滤波, 贝叶斯融合规则, 卡尔曼融合规则, 目标跟踪

Abstract:

To realize multi-sensor tracking for maneuvering target state, this paper presents a multi-sensor fusion algorithm based on DSmT and particle filtering. On the basis of observation date which is delivered by multi-sensor and filtered by particle filters. It chooses DSmT as the combining tool. The observation data is transformed into the elements and masses of the frame of discernment, and gets the finally result. Experimental results show that this algorithm can reduce distance error and improve tracking accuracy, and the appropriate computational complexity can satisfy the demand of fusing on-line.

Key words: DSmT technology, particle filtering, Bayesian fusion rule, Kalman fusion rule, target tracking

中图分类号: