Multi-target tracking using sample-based data association for mixed images

Ting Hao Zhang; Hsiao Tzu Chen; Chih Wei Tang

doi:10.1007/978-3-319-27857-5_12

Multi-target tracking using sample-based data association for mixed images

Ting Hao Zhang, Hsiao Tzu Chen, Chih Wei Tang

Department of Communication Engineering

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution › peer-review

1 Scopus citations

Abstract

The ubiquitous specular reflections arose from glasses seriously degrade accuracy of previous visual trackers. Although there have been few visual tracking schemes developed for mixed images with reflections, none focuses on the issue of multi-target tracking. Thus, this paper proposes a multi-target tracking scheme on mixed images with reflections. In the framework of particle filter, the proposed scheme combines the sample-based joint probabilistic data association filter (SJPDAF) with a single target based tracker that uses co-inference and maximum likelihood for visual cue integration to improve tracking accuracy of multiple targets. The co-inference predicted states are used for measurement validation of the SJPDAF. Experimental results show that the proposed scheme works well compared with the SJPDAF tracker.

Original language	English
Title of host publication	Advances in Visual Computing - 11th International Symposium, ISVC 2015, Proceedings
Editors	Mark Elendt, Richard Boyle, Eric Ragan, Bahram Parvin, Rogerio Feris, Tim McGraw, Ioannis Pavlidis, Regis Kopper, George Bebis, Darko Koracin, Zhao Ye, Gunther Weber
Publisher	Springer Verlag
Pages	127-137
Number of pages	11
ISBN (Print)	9783319278568
DOIs	https://doi.org/10.1007/978-3-319-27857-5_12
State	Published - 2015
Event	11th International Symposium on Advances in Visual Computing, ISVC 2015 - Las Vegas, United States Duration: 14 Dec 2015 → 16 Dec 2015

Publication series

Name	Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume	9474
ISSN (Print)	0302-9743
ISSN (Electronic)	1611-3349

Conference

Conference	11th International Symposium on Advances in Visual Computing, ISVC 2015
Country/Territory	United States
City	Las Vegas
Period	14/12/15 → 16/12/15

Access to Document

10.1007/978-3-319-27857-5_12

反射混合影像序列之多目標追蹤技術研究
Tang, C.-W. (PI)
1/08/15 → 31/07/16
Project: Research

Cite this

Zhang, T. H., Chen, H. T., & Tang, C. W. (2015). Multi-target tracking using sample-based data association for mixed images. In M. Elendt, R. Boyle, E. Ragan, B. Parvin, R. Feris, T. McGraw, I. Pavlidis, R. Kopper, G. Bebis, D. Koracin, Z. Ye, & G. Weber (Eds.), Advances in Visual Computing - 11th International Symposium, ISVC 2015, Proceedings (pp. 127-137). (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 9474). Springer Verlag. https://doi.org/10.1007/978-3-319-27857-5_12

Zhang, Ting Hao ; Chen, Hsiao Tzu ; Tang, Chih Wei. / Multi-target tracking using sample-based data association for mixed images. Advances in Visual Computing - 11th International Symposium, ISVC 2015, Proceedings. editor / Mark Elendt ; Richard Boyle ; Eric Ragan ; Bahram Parvin ; Rogerio Feris ; Tim McGraw ; Ioannis Pavlidis ; Regis Kopper ; George Bebis ; Darko Koracin ; Zhao Ye ; Gunther Weber. Springer Verlag, 2015. pp. 127-137 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)).

@inproceedings{3b79a833dc8548178b2118ea9df73988,

title = "Multi-target tracking using sample-based data association for mixed images",

abstract = "The ubiquitous specular reflections arose from glasses seriously degrade accuracy of previous visual trackers. Although there have been few visual tracking schemes developed for mixed images with reflections, none focuses on the issue of multi-target tracking. Thus, this paper proposes a multi-target tracking scheme on mixed images with reflections. In the framework of particle filter, the proposed scheme combines the sample-based joint probabilistic data association filter (SJPDAF) with a single target based tracker that uses co-inference and maximum likelihood for visual cue integration to improve tracking accuracy of multiple targets. The co-inference predicted states are used for measurement validation of the SJPDAF. Experimental results show that the proposed scheme works well compared with the SJPDAF tracker.",

author = "Zhang, {Ting Hao} and Chen, {Hsiao Tzu} and Tang, {Chih Wei}",

note = "Publisher Copyright: {\textcopyright} Springer International Publishing Switzerland 2015.; 11th International Symposium on Advances in Visual Computing, ISVC 2015 ; Conference date: 14-12-2015 Through 16-12-2015",

year = "2015",

doi = "10.1007/978-3-319-27857-5_12",

language = "???core.languages.en_GB???",

isbn = "9783319278568",

series = "Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)",

publisher = "Springer Verlag",

pages = "127--137",

editor = "Mark Elendt and Richard Boyle and Eric Ragan and Bahram Parvin and Rogerio Feris and Tim McGraw and Ioannis Pavlidis and Regis Kopper and George Bebis and Darko Koracin and Zhao Ye and Gunther Weber",

booktitle = "Advances in Visual Computing - 11th International Symposium, ISVC 2015, Proceedings",

}

Zhang, TH, Chen, HT & Tang, CW 2015, Multi-target tracking using sample-based data association for mixed images. in M Elendt, R Boyle, E Ragan, B Parvin, R Feris, T McGraw, I Pavlidis, R Kopper, G Bebis, D Koracin, Z Ye & G Weber (eds), Advances in Visual Computing - 11th International Symposium, ISVC 2015, Proceedings. Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), vol. 9474, Springer Verlag, pp. 127-137, 11th International Symposium on Advances in Visual Computing, ISVC 2015, Las Vegas, United States, 14/12/15. https://doi.org/10.1007/978-3-319-27857-5_12

Multi-target tracking using sample-based data association for mixed images. / Zhang, Ting Hao; Chen, Hsiao Tzu; Tang, Chih Wei.
Advances in Visual Computing - 11th International Symposium, ISVC 2015, Proceedings. ed. / Mark Elendt; Richard Boyle; Eric Ragan; Bahram Parvin; Rogerio Feris; Tim McGraw; Ioannis Pavlidis; Regis Kopper; George Bebis; Darko Koracin; Zhao Ye; Gunther Weber. Springer Verlag, 2015. p. 127-137 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 9474).

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution › peer-review

TY - GEN

T1 - Multi-target tracking using sample-based data association for mixed images

AU - Zhang, Ting Hao

AU - Chen, Hsiao Tzu

AU - Tang, Chih Wei

N1 - Publisher Copyright: © Springer International Publishing Switzerland 2015.

PY - 2015

Y1 - 2015

N2 - The ubiquitous specular reflections arose from glasses seriously degrade accuracy of previous visual trackers. Although there have been few visual tracking schemes developed for mixed images with reflections, none focuses on the issue of multi-target tracking. Thus, this paper proposes a multi-target tracking scheme on mixed images with reflections. In the framework of particle filter, the proposed scheme combines the sample-based joint probabilistic data association filter (SJPDAF) with a single target based tracker that uses co-inference and maximum likelihood for visual cue integration to improve tracking accuracy of multiple targets. The co-inference predicted states are used for measurement validation of the SJPDAF. Experimental results show that the proposed scheme works well compared with the SJPDAF tracker.

AB - The ubiquitous specular reflections arose from glasses seriously degrade accuracy of previous visual trackers. Although there have been few visual tracking schemes developed for mixed images with reflections, none focuses on the issue of multi-target tracking. Thus, this paper proposes a multi-target tracking scheme on mixed images with reflections. In the framework of particle filter, the proposed scheme combines the sample-based joint probabilistic data association filter (SJPDAF) with a single target based tracker that uses co-inference and maximum likelihood for visual cue integration to improve tracking accuracy of multiple targets. The co-inference predicted states are used for measurement validation of the SJPDAF. Experimental results show that the proposed scheme works well compared with the SJPDAF tracker.

UR - http://www.scopus.com/inward/record.url?scp=84952665335&partnerID=8YFLogxK

U2 - 10.1007/978-3-319-27857-5_12

DO - 10.1007/978-3-319-27857-5_12

M3 - 會議論文篇章

AN - SCOPUS:84952665335

SN - 9783319278568

T3 - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)

SP - 127

EP - 137

BT - Advances in Visual Computing - 11th International Symposium, ISVC 2015, Proceedings

A2 - Elendt, Mark

A2 - Boyle, Richard

A2 - Ragan, Eric

A2 - Parvin, Bahram

A2 - Feris, Rogerio

A2 - McGraw, Tim

A2 - Pavlidis, Ioannis

A2 - Kopper, Regis

A2 - Bebis, George

A2 - Koracin, Darko

A2 - Ye, Zhao

A2 - Weber, Gunther

PB - Springer Verlag

T2 - 11th International Symposium on Advances in Visual Computing, ISVC 2015

Y2 - 14 December 2015 through 16 December 2015

ER -

Zhang TH, Chen HT, Tang CW. Multi-target tracking using sample-based data association for mixed images. In Elendt M, Boyle R, Ragan E, Parvin B, Feris R, McGraw T, Pavlidis I, Kopper R, Bebis G, Koracin D, Ye Z, Weber G, editors, Advances in Visual Computing - 11th International Symposium, ISVC 2015, Proceedings. Springer Verlag. 2015. p. 127-137. (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)). doi: 10.1007/978-3-319-27857-5_12

Multi-target tracking using sample-based data association for mixed images

Abstract

Publication series

Conference

Access to Document

Fingerprint

Projects

反射混合影像序列之多目標追蹤技術研究

Cite this