Robust visual tracking based on structured sparse representation model

• Published in: Multimedia tools and applications Vol. 74; no. 3; pp. 1021 - 1043
• Main Authors: Zhang, Hanling, Tao, Fei, Yang, Gaobo
• Format: Journal Article
• Language: English
• Published: Boston Springer US 01.02.2015; Springer Nature B.V
• Subjects: Algorithms; Analysis; Computer Communication Networks; Computer Science; Data Structures and Information Theory; Mathematical models; Methods; Multimedia computer applications; Multimedia Information Systems; Occlusion; Reconstruction; Representations; Special Purpose and Application-Based Systems; State of the art; Studies; Tracking; Tracking control systems; Vision systems; Visual; Sparse representation; Block division; Visual tracking; Particle filter; Template update
• ISSN: 1380-7501; 1573-7721
• DOI: 10.1007/s11042-013-1709-0

Sparse representation has been one of the most influential frameworks for visual tracking. However, most tracking methods based on sparse representation only consider the holistic representation and lack local information, which may lead to fail when there is similar object or occlusion in the scene. In this paper, we present a novel robust visual tracking algorithm based on structured sparse representation model. This model includes one fixed template, nine variational templates and the background templates, which are selectively updated to adapt to the appearance change of the target. And the update scheme is developed by exploiting the strength of the incremental PCA learning and sparse representation. By incorporating the block-division feature into sparse representation framework, it can capture the intrinsic structured distribution of sparse coefficients effectively and reduce the influence of the occluded target template. In addition, we propose a sparsity-based discriminative classifier, which employ the distinction of reconstruction error between the foreground and the background to improve discrimination performance for object tracking. Both qualitative and quantitative evaluations on benchmark challenging sequences demonstrate that the proposed tracking algorithm performs favorably against several state-of-the-art tracking methods.