8 April 2013 One-class support vector machine-assisted robust tracking
Author Affiliations +
J. of Electronic Imaging, 22(2), 023002 (2013). doi:10.1117/1.JEI.22.2.023002
Recently, tracking is regarded as a binary classification problem by discriminative tracking methods. However, such binary classification may not fully handle the outliers, which may cause drifting. We argue that tracking may be regarded as one-class problem, which avoids gathering limited negative samples for background description. Inspired by the fact the positive feature space generated by one-class support vector machine (SVM) is bounded by a closed hyper sphere, we propose a tracking method utilizing one-class SVMs that adopt histograms of oriented gradient and 2bit binary patterns as features. Thus, it is called the one-class SVM tracker (OCST). Simultaneously, an efficient initialization and online updating scheme is proposed. Extensive experimental results prove that OCST outperforms some state-of-the-art discriminative tracking methods that tackle the problem using binary classifiers on providing accurate tracking and alleviating serious drifting.
© 2013 SPIE and IS&T
Keren Fu, Chen Gong, Yu Qiao, Jie Yang, Irene Yu-Hua Gu, "One-class support vector machine-assisted robust tracking," Journal of Electronic Imaging 22(2), 023002 (8 April 2013). http://dx.doi.org/10.1117/1.JEI.22.2.023002

Binary data

Optical spheres

Detection and tracking algorithms

Feature extraction



Zoom lenses


Deblocking of mobile stereo video
Proceedings of SPIE (February 21 2012)
Effects Of Rectification On Image Correlation
Proceedings of SPIE (December 23 1980)
Video coding mode decision as a classification problem
Proceedings of SPIE (January 18 2010)
Path Planning Using Potential Field Representation
Proceedings of SPIE (March 29 1988)

Back to Top