The Community for Technology Leaders
Green Image
ISSN: 0162-8828
Junghyun Kwon , TeleSecurity Sciences, Inc., Las Vegas
Hee Seok Lee , Seoul National University, Seoul
Frank C. Park , Seoul National University, Seoul
Kyoung Mu Lee , Seoul National University, Seoul
Existing approaches to template-based visual tracking, in which the objective is to continuously estimate the spatial transformation parameters of an object template over video frames, have primarily been based on deterministic optimization, which as is well-known can result in convergence to local optima. To overcome this limitation of the deterministic optimization approach, in this paper we present a novel particle filtering approach to template-based visual tracking. We formulate the problem as a particle filtering problem on matrix Lie groups, specifically the Special Linear group SL(3) and the two-dimensional affine group Aff (2). Computational performance and robustness are enhanced through a number of features: (i) Gaussian importance functions on the groups are iteratively constructed via local linearization; (ii) the inverse formulation of the Jacobian calculation is used; (iii) template resizing is performed; and (iv) parent-child particles are developed and used. Extensive experimental results using challenging video sequences demonstrate the enhanced performance and robustness of our particle filtering-based approach to template-based visual tracking. We also show that our approach outperforms several state-of-the-art template-based visual tracking methods via experiments using the publicly available benchmark dataset.
Tracking, Visualization, Equations, Mathematical model, Approximation methods, Algebra, Approximation algorithms, Gaussian importance function, visual tracking, object template, particle filtering, Lie group, special linear group, affine group

F. C. Park, H. S. Lee, J. Kwon and K. M. Lee, "A Geometric Particle Filter for Template-Based Visual Tracking," in IEEE Transactions on Pattern Analysis & Machine Intelligence.
175 ms
(Ver 3.3 (11022016))