基于多个颜色分布模型的Camshift跟踪算法

被引:30
作者
左军毅
梁彦
潘泉
赵春晖
张洪才
机构
[1] 西北工业大学自动化学院
基金
国家自然科学基金重点项目;
关键词
目标跟踪; 概率图; 均值漂移; 多模型;
D O I
暂无
中图分类号
TP391.41 [];
学科分类号
080203 ;
摘要
视角变化往往会引起目标外观特征的变化,传统的基于单一颜色直方图模型的Camshift跟踪算法往往不能适应这种变化.为此,本文从提高模型描述能力入于,提出利用目标外观的先验知识,为目标建立多个颜色模型,在此基础上设计目标函数,通过对目标函数的优化,实时地为每帧跟踪从多个模型的凸组合中选取最优模型.另外,在对Camshift算法深入研究的过程中,发现了概率图平均亮度和图像块颜色分布之间的一种定量关系,这种关系为进一步理解多模型算法的工作机理提供了帮助.头部跟踪的实验结果表明,与单一固定模型以及自适应单模型算法相比,多模型Camshift算法对目标外观的快速变化适应性很强,而且计算代价不大.
引用
收藏
页码:736 / 742
页数:7
相关论文
共 10 条
[1]  
Robust tracking with mo-tion esti mation and local kernel-based color modeling. Babu R V,Perez P,Bouthemy P. Image and Vision Computing . 2007
[2]  
Mean Shift:a robust approach to- ward feature space analysis. Comaniciu D,Meer P. IEEE Transactions on Pattern Analysis and Machine Intelligence . 2002
[3]  
Adaptive probabilistic visual tracking with incremental subspace update. Ross D,Lim J,Yang M H. Proceed- ings of the 8th European Conference on Computer Vision . 2004
[4]  
Real time face and object tracking as a com- ponent of a perceptual user interface. Bradski G R. Proceedings of the 4th Workshop on Application of Computer Vision . 1998
[5]  
Kernel-based object tracking. Comaniciu D,Ramesh V,Meer P. IEEE Transactions on Pattern Analysis and Ma-chine Intelligence . 2003
[6]  
The template update problem. Matthews L,Ishikawa T,Baker S. IEEE Transactions on Pattern Analysis and Ma- chine Intelligence . 2004
[7]  
Spatio-temporal context for robust multitarget tracking. Nguyen H T,Ji Q,Smeulders A W M. IEEE Transactions on Pattern Analysis and Machine Intelligence . 2007
[8]  
Online learning of probabilistic ap- pearance manifolds for video-based recognition and track- ing. Lee K C,Kriegman D. Proceedings of IEEE Computer Society Conference on Computer Vision and Pattern Recognition . 2005
[9]  
Visual tracking and recognition using appearance-adaptive models in particle filters. S. H. K. Zhou,R. Chellappa,and B. Moghaddam. IEEE Transactions on Image Processing . 2004
[10]  
Online selection of discriminative tracking features. Collins R T,Liu Y X,Leordeanu M. IEEE Transactions on Pattern AnaJysis and Machine Intelligence . 2005