A modified KLT multiple objects tracking framework based on global segmentation and adaptive template

Author

Kang Xue ; Vela, Patricio A. ; Yue Liu ; Yongtian Wang

Author_Institution

Beijing Inst. of Technol., Beijing, China

fYear

2012

fDate

11-15 Nov. 2012

Firstpage

3561

Lastpage

3564

Abstract

This paper presents a modified Kanade-Lucas-Tomasi (KLT) tracking framework for multiple objects tracking applications. First, the framework includes a global pixel-level probabilistic model and an adaptive RGB template model to modify traditional KLT tracker more robust to track multiple objects and partial occlusions. Meanwhile, a Merge and Split algorithm is introduced in the proposed framework to track complete occlusions. The advantage of our method is demonstrated on a variety of challenging video sequences.

Keywords

hidden feature removal; image colour analysis; image segmentation; image sequences; object tracking; probability; KLT tracker; Kanade-Lucas-Tomasi tracking framework; adaptive RGB template model; adaptive template; global pixel-level probabilistic model; global segmentation; merge and split algorithm; modified KLT multiple object tracking framework; partial occlusions; video sequences; Adaptation models; Mathematical model; Object tracking; Principal component analysis; Probabilistic logic; Robustness; Target tracking;

fLanguage

English

Publisher

ieee

Conference_Titel

Pattern Recognition (ICPR), 2012 21st International Conference on

Conference_Location

Tsukuba

ISSN

1051-4651

Print_ISBN

978-1-4673-2216-4

Type

conf

Filename

6460934