计算机应用 ›› 2011, Vol. 31 ›› Issue (10): 2731-2733.DOI: 10.3724/SP.J.1087.2011.02731

• 图形图像技术 • 上一篇    下一篇

基于多线索自适应融合的抗遮挡目标跟踪算法

曹洁,付德强   

  1. 兰州理工大学 电气工程与信息工程学院, 兰州 730050
  • 收稿日期:2011-04-25 修回日期:2011-06-17 发布日期:2011-10-11 出版日期:2011-10-01
  • 通讯作者: 付德强
  • 作者简介:曹洁(1966-),女,安徽宿州人,教授,博士生导师,主要研究方向:信息融合、智能交通;付德强(1983-),男,山东邹城人,硕士研究生,主要研究方向:信息融合、目标跟踪。
  • 基金资助:

    甘肃省自然科学基金资助项目(1010RJZA046);甘肃省教育厅研究生导师科研项目(0914ZTB003)

Anti-occlusion object tracking algorithm based on adaptive multiple cues fusion

CAO Jie,FU De-qiang   

  1. College of Electrical and Information Engineering, Lanzhou University of Technology, Lanzhou Gansu 730050, China
  • Received:2011-04-25 Revised:2011-06-17 Online:2011-10-11 Published:2011-10-01

摘要: 针对在复杂背景下,特别当目标与背景颜色相似、目标被遮挡时容易导致跟踪失败的问题,采用颜色与纹理两种互补特征融合的粒子滤波算法,同时提出一种融合策略自适应的抗遮挡跟踪方法,当遮挡发生时,适时切换融合策略,并在粒子滤波框架内嵌入mean-shift算法,克服了粒子退化现象。实验结果表明,该算法具有较强的抗遮挡能力,能够跟踪复杂背景下的目标。

关键词: 目标跟踪, 颜色特征, 纹理特征, 自适应融合, 抗遮挡

Abstract: Complex background, especially when the object is similar to the background in color or the target gets blocked, can easily lead to tracking failure. Therefore, a particle filter integrating color feature and texture feature was proposed. An object tracking algorithm based on adaptive multiple cues fusion was proposed too. When occlusion occurred, the fusion rules would be switched in time. Mean-shift algorithm was applied to each particle to overcome degeneracy. The experimental results show that this algorithm is robust to occlusion and is able to track object under complex background.

Key words: object tracking, color feature, texture feature, adaptive fusion, anti-occlusion

中图分类号: