东北大学学报(自然科学版) ›› 2013, Vol. 34 ›› Issue (12): 1678-1681.DOI: 10.12068/j.issn.1005-3026.2013.12.002

• 信息与控制 • 上一篇    下一篇

基于高效二阶最小化的运动模糊目标跟踪算法

赵越,李晶皎,李贞妮,马技   

  1. (东北大学信息科学与工程学院,辽宁沈阳110819)
  • 发布日期:2013-07-09
  • 通讯作者: 赵越
  • 作者简介:赵越(1979-),女,辽宁抚顺人,东北大学博士研究生,渤海大学讲师;李晶皎(1964-),女,辽宁沈阳人,东北大学教授,博士生导师.
  • 基金资助:
    国家自然科学基金资助项目(60970157).

Motion Blur Target Tracking Method Based on Efficient SecondOrder Minimization〓

ZHAO Yue, LI Jingjiao, LI Zhenni, MA Ji   

  1. School of Information Science & Engineering, Northeastern University, Shenyang 110819, China.
  • Published:2013-07-09
  • Contact: ZHAO Yue
  • About author:-
  • Supported by:
    -

摘要: 针对运动模糊严重时容易导致目标跟踪失败的问题,提出了基于高效二阶最小化(ESM)的模板匹配目标跟踪算法.首先提出一种运动模糊模板匹配的图像构造模型.然后引入ESM算法,在ESM算法基础上,用改进的高效二阶最小化(ESM-MB)算法跟踪运动模板.引入摄像头快门估计时间作为参数,提出了自适应不同快门速度所引起的不同的运动模糊的ESM-MB-ST跟踪算法.最后,通过真实视频序列的跟踪实验,验证了提出的ESM-MB算法及ESM-MB-ST算法具有更强的鲁棒性与实时性.

关键词: 高效二阶最小化, 运动模糊, 图像构造模型, 快门时间, 模板匹配

Abstract: As severe motion blur will lead to the failure of target tracking, an ESMbased template matching target tracking algorithm was proposed. First, an image structure model of motionblurred template matching was presented. Then, the ESM algorithm was introduced and motion template was tracked based on ESMMB. Subsequently, camera shutter estimated time was introduced as a parameter and selfadaptive ESMMBST tracking algorithm of motion blur caused by different shutter speed was proposed. Finally, the real video sequence tracking experiments indicated that the proposed ESMMB algorithm and the ESMMBST algorithm had a better robustness and realtime ability.

Key words: efficient secondorder minimization, motion blur, image formation model, shutter time, template matching

中图分类号: