J4 ›› 2012, Vol. 50 ›› Issue (03): 535-.

• 计算机科学 • 上一篇    下一篇

基于Meanshift优化粒子滤波算法的低空平台下车辆跟踪

史红1, 刘涛2, 李鸣3, 沈敏洁4   

  1. 1. 吉林师范大学 信息技术学院, 吉林 四平 136000|2. 吉林大学 计算机科学与技术学院, 长春 130012;3. 吉林省经济信息中心, 长春 130061|
    4. 国家专利局专利审查协作北京中心, |北京 100083
  • 收稿日期:2011-09-13 出版日期:2012-05-26 发布日期:2012-05-28
  • 通讯作者: 史红 E-mail:shihong8498@yahoo.com.cn

Meanshift Optimization Based Particle Filter Tracking of Vehicles in LowAltitude Platform

SHI Hong1, LIU Tao2, LI Ming3, SHEN Min jie4   

  1. 1. College of Information and Technology, Jilin Normal University, Siping 136000, Jilin Province, China;2. College of Computer Science and Technology, Jilin University, Changchun 130012, China;3. Economic Information Center of Jilin Province, Changchun 130061, China;4. Patent Examination Cooperation Center of the Patent Office, Beijing 100083, China
  • Received:2011-09-13 Online:2012-05-26 Published:2012-05-28
  • Contact: SHI Hong E-mail:shihong8498@yahoo.com.cn

摘要:

针对低空平台下运行车辆的特点, 提出一种基于Meanshift粒子优化的粒子滤波算法实现低空平台下的车辆跟踪. 该算法使用颜色表示目标, 通过Meanshift算法对粒子滤波进行迭代优化, 减少了稳健跟踪一个目标所需的粒子数, 提高了算法的运行效率, 在小目标和多目标的情况下也能稳健跟踪. 实验结果表明, 该算法具有较强的鲁棒性和稳定性, 能实现低空平台下目标车辆的快速跟踪.

关键词: 粒子滤波, 低空平台, 跟踪

Abstract:

We analyzed the vehicle characteristics in low altitude platform, and proposed a fast object tracking method via particle filter based on Meanshift optimization. In our method, we used color distribution histogram to represent an object. Then Meanshift algorithm was applied to iterative optimization in particle selecting stage. This innovation is able to reduce the particle number in tracking. The experiments show that our method has strong robustness and stability, and it can be applied in fast tracking in low altitude platform.

Key words: particle filter, lowaltitude platform, tracking

中图分类号: 

  • TP391