Regular Articles

Robust visual ℓ2-regularized least squares tracker with Bayes classifier and coding error

[+] Author Affiliations
Yueen Hou

South China University of Technology, School of Mechanical and Automotive Engineering, Wushan Road 381, Guangzhou, China

Weiguang Li

South China University of Technology, School of Mechanical and Automotive Engineering, Wushan Road 381, Guangzhou, China

Aiqiong Rong

South China University of Technology, School of Mechanical and Automotive Engineering, Wushan Road 381, Guangzhou, China

Huidong Lou

South China University of Technology, School of Mechanical and Automotive Engineering, Wushan Road 381, Guangzhou, China

Sibo Quan

South China University of Technology, School of Mechanical and Automotive Engineering, Wushan Road 381, Guangzhou, China

J. Electron. Imaging. 22(4), 043036 (Dec 20, 2013). doi:10.1117/1.JEI.22.4.043036
History: Received June 5, 2013; Revised October 30, 2013; Accepted November 20, 2013
Text Size: A A A

Abstract.  Visual object tracking is a challenging task due to the existence of significant illumination change, appearance variation, and occlusion. We propose a structured collaborative representation-based visual tracking algorithm by using both the generative appearance model and the discriminative model. First, a structured collaborative representation model based on 2-regularized least squares is used to exploit both holistic and local information of the target. In the structured collaborative representation framework, a new over-complete dictionary containing both target and background templates is proposed to enhance the robustness of the tracking algorithm. Second, the tracking task is treated as a binary classification problem, and a Bayes classifier is trained online by the use of structured collaborative representation coefficients of positive and negative samples. Furthermore, a residual error score is constructed to improve the detective ability of the tracker. Finally, target templates are updated by combining incremental subspace learning and collaborative representation together, and background templates are subsequently updated by samples around latest results. Compared with four state-of-the-art tracking algorithms in 12 challenging video sequences, the proposed tracking algorithm demonstrates a better performance overall in terms of experimental results.

© 2013 SPIE and IS&T

Citation

Yueen Hou ; Weiguang Li ; Aiqiong Rong ; Huidong Lou and Sibo Quan
"Robust visual ℓ2-regularized least squares tracker with Bayes classifier and coding error", J. Electron. Imaging. 22(4), 043036 (Dec 20, 2013). ; http://dx.doi.org/10.1117/1.JEI.22.4.043036


Access This Article
Sign in or Create a personal account to Buy this article ($20 for members, $25 for non-members).

Some tools below are only available to our subscribers or users with an online account.

Related Content

Customize your page view by dragging & repositioning the boxes below.

Related Book Chapters

Topic Collections

Advertisement
  • Don't have an account?
  • Subscribe to the SPIE Digital Library
  • Create a FREE account to sign up for Digital Library content alerts and gain access to institutional subscriptions remotely.
Access This Article
Sign in or Create a personal account to Buy this article ($20 for members, $25 for non-members).
Access This Proceeding
Sign in or Create a personal account to Buy this article ($15 for members, $18 for non-members).
Access This Chapter

Access to SPIE eBooks is limited to subscribing institutions and is not available as part of a personal subscription. Print or electronic versions of individual SPIE books may be purchased via SPIE.org.