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

通訊工程學系

研究成果: 書貢獻/報告類型 › 會議論文篇章 › 同行評審

1 引文斯高帕斯（Scopus）

摘要

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.

原文	???core.languages.en_GB???
主出版物標題	Advances in Visual Computing - 11th International Symposium, ISVC 2015, Proceedings
編輯	Mark Elendt, Richard Boyle, Eric Ragan, Bahram Parvin, Rogerio Feris, Tim McGraw, Ioannis Pavlidis, Regis Kopper, George Bebis, Darko Koracin, Zhao Ye, Gunther Weber
發行者	Springer Verlag
頁面	127-137
頁數	11
ISBN（列印）	9783319278568
DOIs	https://doi.org/10.1007/978-3-319-27857-5_12
出版狀態	已出版 - 2015
事件	11th International Symposium on Advances in Visual Computing, ISVC 2015 - Las Vegas, United States 持續時間: 14 12月 2015 → 16 12月 2015

出版系列

名字	Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
卷	9474
ISSN（列印）	0302-9743
ISSN（電子）	1611-3349

???event.eventtypes.event.conference???

???event.eventtypes.event.conference???	11th International Symposium on Advances in Visual Computing, ISVC 2015
國家/地區	United States
城市	Las Vegas
期間	14/12/15 → 16/12/15

存取文件

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
研究計畫: Research

引用此

Zhang, T. H., Chen, H. T., & Tang, C. W. (2015). Multi-target tracking using sample-based data association for mixed images. 於 M. Elendt, R. Boyle, E. Ragan, B. Parvin, R. Feris, T. McGraw, I. Pavlidis, R. Kopper, G. Bebis, D. Koracin, Z. Ye, & G. Weber (編輯), Advances in Visual Computing - 11th International Symposium, ISVC 2015, Proceedings (頁 127-137). (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); 卷 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. 編輯 / 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. 頁 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. 於 M Elendt, R Boyle, E Ragan, B Parvin, R Feris, T McGraw, I Pavlidis, R Kopper, G Bebis, D Koracin, Z Ye & G Weber (編輯), 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), 卷 9474, Springer Verlag, 頁 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. 編輯 / 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); 卷 9474).

研究成果: 書貢獻/報告類型 › 會議論文篇章 › 同行評審

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. 於 Elendt M, Boyle R, Ragan E, Parvin B, Feris R, McGraw T, Pavlidis I, Kopper R, Bebis G, Koracin D, Ye Z, Weber G, 編輯, 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

摘要

出版系列

???event.eventtypes.event.conference???

存取文件

指紋

專案

反射混合影像序列之多目標追蹤技術研究

引用此