基于自适应分块的视频目标跟踪方法研究倡.pdf

文档介绍：第 27 卷第 9 .9
2010 年 9 月 Vol .2010No
Application Research puters Sep
基于自适应分块的视频目标跟踪方法研究倡
郝慧娟, 许继勇, 刘广起
(山东省计算中心, 济南 250014)
摘要: 针对目标跟踪中的遮挡问题,提出了一种基于颜色特征的自适应跟踪算法。该算法利用模糊均值聚
K
类将目标自适应分块,采用单调递减的核函数对子块进行加权,目标模型的更新只需更新子块的颜色均值,计算
量小。在跟踪过程采用目标整体匹配的方法,结合设计的自适应滤波器,有效地解决了跟踪过程中的目
Kalman
标遮挡问题。实验结果表明,新算法能够实现目标的准确跟踪。
关键词: 目标跟踪; 颜色直方图; 自适应分块; 滤波
Kalman
中图分类号: 文献标志码: 文章编号: 1001唱3695(2010)09唱3579唱02
TP A
: /. .
doi j issn
Study on vision object tracking based on adaptive object segmentation
唱, 唱, 唱
HAO Hui juan XU Ji yong LIU Guang qi
( puter Science Center, Jinan 250014,China)
Abstract: ,
In order. to solve the object tracking under occlusion唱 this paper proposed the, adaptively tracking algorithm唱 based
on color features Adaptively divided the. object using fuzzy K means clustering algorithm and weighted唱 with’ the sub,regions
the monotone decreasing. kernel function It updated, The object model through the mean, value of sub regions color so the
calculation, was simple During the object tracking used the method. of integral bining with the adaptive Kalman唱
filter resolved. the object tracking under occlusion effectively The experiments show that the new algorithm can track the ob
jectKey exactly words: ; ; ;
object tracking color histogram adaptive object segmentation Kalman filtering
引言