计算机工程与应用 ›› 2009, Vol. 45 ›› Issue (28): 177-179.DOI: 10.3778/j.issn.1002-8331.2009.28.053

• 图形、图像、模式识别 • 上一篇    下一篇

CAMSHIFT与基于核的目标跟踪算法的比较与分析

宁纪锋1,2,姜 光2,吴成柯2   

  1. 1.西北农林科技大学 信息工程学院,陕西 杨凌 712100
    2.西安电子科技大学 综合业务网国家重点实验室,西安 710071
  • 收稿日期:2008-12-29 修回日期:2009-02-13 出版日期:2009-10-01 发布日期:2009-10-01
  • 通讯作者: 宁纪锋

Comparison and analysis of CAMSHIFT and kernel based on object tracking

NING Ji-feng1,2,JIANG Guang2,WU Cheng-ke2   

  1. 1.College of Information Engineering,Northwest A&F University,Yangling,Shaanxi 712100,China
    2.State Key Laboratory of ISN,Xidian University,Xi’an 710071,China
  • Received:2008-12-29 Revised:2009-02-13 Online:2009-10-01 Published:2009-10-01
  • Contact: NING Ji-feng

摘要: CAMSHIFT和基于核的目标跟踪是两种经典的基于Mean Shift的目标跟踪算法,它们的实现过程有许多类似之处。为了说明在实际应用中如何选择合理的跟踪方案,从目标模型、候选模型、核函数、迭代过程等方面对二者进行了深入的比较和分析,指出了二者的特点和区别,对于正确理解和使用这两种方法将会有一定的帮助。

关键词: Mean Shift, 连续自适应均值移动跟踪算法(CAMSHIFT), 基于核的目标跟踪, 权值图像

Abstract: CAMSHIFT and kernel based object tracking are two classical tracking models based on mean shift algorithm,which are very similar in many ways.In order to guide to select the proper tracking method in the real application,this paper analyzes them and points the difference between them from target model,candidate model and iteration equation.The discussion on them will provide a useful help for the proper understanding and use of these two tracking models.

Key words: Mean Shift, Continuously Adaptive Mean Shift(CAMSHIFT), kernel based object tracking, weight image

中图分类号: