Robust Tracking Based on Improved Mean-shift and Hybrid Approach

被引：0
|
作者：
Chen, Liangshi [1 ]
Wu, Juan [2 ]
Pang, Tao [2 ]
机构：
[1] Sun Yat Sen Univ, Sch Software, Guangzhou 510006, Guangdong, Peoples R China
[2] ChinaTelecom Co Ltd, Guangdong Res Inst, Publ Customer Prod R&D Dept, Guangzhou 510630, Guangdong, Peoples R China
来源：
2011 INTERNATIONAL CONFERENCE ON COMPUTER, COMMUNICATION AND INFORMATION TECHNOLOGY (ICCCIT 2011) | 2011年
关键词：
Object Tracking; Mean Shift; Spatial-color Histogram; Kalman Filter; Particle Filter;
D O I：
暂无
中图分类号：
TP [自动化技术、计算机技术];
学科分类号：
0812 ;
摘要：
A simple but useful and robust tracking algorithm is proposed. Combining efficiency and reliability is our major objective. The main contribution of this paper is to improve and to combine tracking algorithms to obtain a hybrid tracking method which achieves better tracking effect intuitively, particularly under occlusions. A major disadvantage of mean shift is losing the layout information of targets' colors. Mean shift is modified aiming at this drawback. After analyzing pros and cons of several methods, a hybrid tracking algorithm which relies mainly on the improved mean shift, assisted by Kalman filter and particle filter is proposed. Experiments show that tracking accuracy is effectively improved.
引用
收藏
页码：134 / 138
页数：5
相关论文
共 15 条
  • [1] [Anonymous], 2008, PROC IEEE C COMPUTER
  • [2] [Anonymous], P 11 INT IEEE C INT
  • [3] Ensemble tracking
    Avidan, Shai
    [J]. IEEE TRANSACTIONS ON PATTERN ANALYSIS AND MACHINE INTELLIGENCE, 2007, 29 (02) : 261 - 271
  • [4] Babenko B, 2009, PROC CVPR IEEE, P983, DOI 10.1109/CVPRW.2009.5206737
  • [5] Birchfield ST, 2005, PROC CVPR IEEE, P1158
  • [6] Mean shift: A robust approach toward feature space analysis
    Comaniciu, D
    Meer, P
    [J]. IEEE TRANSACTIONS ON PATTERN ANALYSIS AND MACHINE INTELLIGENCE, 2002, 24 (05) : 603 - 619
  • [7] Comaniciu D, 2000, PROC CVPR IEEE, P142, DOI 10.1109/CVPR.2000.854761
  • [8] Comaniciu D., 2002, IEEE T PATTERN ANAL, V25, p603 619
  • [9] Dalal N., P 2005 IEEE COMP SOC
  • [10] Felzenszwalb P. F., 2010, IEEE transactions on pattern analysis and machine intelligence, V32, P1627, DOI [DOI 10.1109/TPAMI.2009.167, 10.1109/TPAMI.2009.167]
12