CamShift Guided Unscented Particle Filter for Image-based Object Tracking

被引：0

作者：

Peerapong, T. ^{[1
,3
]}

Usman, M. ^{[2
,3
]}

Gao, Peng-qi ^{[1
,3
]}

Shen, Ming ^{[1
,3
]}

Guo, Xiao-zhang ^{[1
,3
]}

Yang, Da-tao ^{[1
,3
]}

Yu, Huan-huan ^{[1
,3
]}

Zhou, Wei-ping ^{[1
,3
]}

Sun, Ming-guo ^{[1
,3
]}

Wang, Wei-qiang ^{[2
,3
]}

Zhao, You ^{[1
,3
]}

机构：

[1] Chinese Acad Sci NAOC, Natl Astron Observ, Beijing, Peoples R China

[2] Univ Chinese Acad Sci, Sch Comp & Control Engn, Beijing, Peoples R China

[3] Univ Chinese Acad Sci, Beijing, Peoples R China

来源：

INTERNATIONAL CONFERENCE ON INFORMATICS, CONTROL AND AUTOMATION (ICA 2015) | 2015年

关键词：

CamShift; Unscented particle filter; Object tracking; Video surveillance systems;

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：

The Unscented Particle Filter and CamShift (Continuously adaptive mean shift) are two successful approaches for the object tracking in video surveillance systems. We focus on the tracking performance of both algorithms and explore more intrinsic relationship between CamShift and Unscented Particle Filter (CAMSGUPF). Our proposed algorithm provides better sampling by shifting samples to their neighboring modes, passing over the degeneracy problem, and involves fewer particles to maintain multiple hypotheses. We used the Root Mean Square Error (RMSE) to compare our algorithm with traditional "CamShift guided particle filter". Experimental results prove that our proposed algorithm has better performance than traditional method according to the degree of applicability.

引用

页码：246 / 251

页数：6

共 8 条

[1]

Allen J. G., 2004, ACM VIP

[2]

Bradski Gary R, 1998, P IEEE 4 WORKSH APPL

[3]

HAMMERSLEY JM, J ROYAL STAT SOC B, V16, P711

[4]

Isard M., 1996, Proc. Europe Conference on Computer Vision, P343

[5]

Lien S F, 2015, Applied Mathematics & Information Sciences, V9, P395

[6]

Rui Y., 2001, COMP VIS PATT REC 20, V2

[7]

VANDERMERWE R, 2000, UNSCENTED PARTICLE F

[8] CamShift guided particle filter for visual tracking [J].

Wang, Zhaowen ;

Yang, Xiaokang ;

Xu, Yi ;

Yu, Songyu .

PATTERN RECOGNITION LETTERS, 2009, 30 (04) :407-413

← 1 →