A multi-camera tracking system that can always select a better view to perform tracking

Shih Wei Sun; Hsing Yuan Lo; Hong Ju Lin; Yong-Sheng Chen; Fay Huang; Hong Yuan Mark Liao

A multi-camera tracking system that can always select a better view to perform tracking

Shih Wei Sun^*, Hsing Yuan Lo, Hong Ju Lin, Yong-Sheng Chen, Fay Huang, Hong Yuan Mark Liao

^*Corresponding author for this work

Department of Computer Science

Research output: Contribution to conference › Paper › peer-review

4 Scopus citations

Abstract

In this paper, we propose a new multiple-camera people tracking system that is equipped with the following functions: (1) can handle long-term occlusions, complete occlusions, and unpredictable motions; (2) can detect arbitrary sized foreground objects; (3) can detect objects with much faster speed. The main contribution of our method is twofold: 1) An Mto-one relationship with only point homography matching for occlusion detection can achieve efficiency; 2) A view-hopping technique based on object motion probability (OMP) is proposed to automatically select an appropriate observation view for tracking a human subject.

Original language	English
Pages	373-379
Number of pages	7
State	Published - 4 Oct 2009
Event	Asia-Pacific Signal and Information Processing Association 2009 Annual Summit and Conference, APSIPA ASC 2009 - Sapporo, Japan Duration: 4 Oct 2009 → 7 Oct 2009

Conference

Conference	Asia-Pacific Signal and Information Processing Association 2009 Annual Summit and Conference, APSIPA ASC 2009
Country/Territory	Japan
City	Sapporo
Period	4/10/09 → 7/10/09

Cite this

@conference{d27c9dd42ddb4c47b8ed642b7744d5ba,

title = "A multi-camera tracking system that can always select a better view to perform tracking",

abstract = "In this paper, we propose a new multiple-camera people tracking system that is equipped with the following functions: (1) can handle long-term occlusions, complete occlusions, and unpredictable motions; (2) can detect arbitrary sized foreground objects; (3) can detect objects with much faster speed. The main contribution of our method is twofold: 1) An Mto-one relationship with only point homography matching for occlusion detection can achieve efficiency; 2) A view-hopping technique based on object motion probability (OMP) is proposed to automatically select an appropriate observation view for tracking a human subject.",

author = "Sun, {Shih Wei} and Lo, {Hsing Yuan} and Lin, {Hong Ju} and Yong-Sheng Chen and Fay Huang and Liao, {Hong Yuan Mark}",

year = "2009",

month = oct,

day = "4",

language = "English",

pages = "373--379",

note = "null ; Conference date: 04-10-2009 Through 07-10-2009",

}

A multi-camera tracking system that can always select a better view to perform tracking. / Sun, Shih Wei; Lo, Hsing Yuan; Lin, Hong Ju et al.

2009. 373-379 Paper presented at Asia-Pacific Signal and Information Processing Association 2009 Annual Summit and Conference, APSIPA ASC 2009, Sapporo, Japan.

Research output: Contribution to conference › Paper › peer-review

TY - CONF

T1 - A multi-camera tracking system that can always select a better view to perform tracking

AU - Sun, Shih Wei

AU - Lo, Hsing Yuan

AU - Lin, Hong Ju

AU - Chen, Yong-Sheng

AU - Huang, Fay

AU - Liao, Hong Yuan Mark

PY - 2009/10/4

Y1 - 2009/10/4

N2 - In this paper, we propose a new multiple-camera people tracking system that is equipped with the following functions: (1) can handle long-term occlusions, complete occlusions, and unpredictable motions; (2) can detect arbitrary sized foreground objects; (3) can detect objects with much faster speed. The main contribution of our method is twofold: 1) An Mto-one relationship with only point homography matching for occlusion detection can achieve efficiency; 2) A view-hopping technique based on object motion probability (OMP) is proposed to automatically select an appropriate observation view for tracking a human subject.

AB - In this paper, we propose a new multiple-camera people tracking system that is equipped with the following functions: (1) can handle long-term occlusions, complete occlusions, and unpredictable motions; (2) can detect arbitrary sized foreground objects; (3) can detect objects with much faster speed. The main contribution of our method is twofold: 1) An Mto-one relationship with only point homography matching for occlusion detection can achieve efficiency; 2) A view-hopping technique based on object motion probability (OMP) is proposed to automatically select an appropriate observation view for tracking a human subject.

UR - http://www.scopus.com/inward/record.url?scp=77950577896&partnerID=8YFLogxK

M3 - Paper

AN - SCOPUS:77950577896

SP - 373

EP - 379

Y2 - 4 October 2009 through 7 October 2009

ER -

A multi-camera tracking system that can always select a better view to perform tracking

Abstract

Conference

Other files and links

Fingerprint

Cite this