Visual tracking via Boolean map representations. (September 2018)

• Record Type: Journal Article
• Authors: Zhang, Kaihua; Liu, Qingshan; Yang, Jian; Yang, Ming-Hsuan
• Is Part Of: Pattern recognition. Volume 81(2018:Sep.)
• Page Start: 147
• Page End: 160
• Publication Date: 2018-09
• Subjects: Visual tracking -- Boolean map -- Logistic regression -- Intersection kernel; Pattern perception -- Periodicals; Perception des structures -- Périodiques; Patroonherkenning; 006.4
• DOI: 10.1016/j.patcog.2018.03.029
• Languages: English
• ISSNs: 0031-3203

Abstract

Highlights: We demonstrate that the connectivity cues can be effectively used for robust visual tracking. We show that the BMR scheme can be approximated as an explicit feature map of the intersection kernel. Our tracking algorithm based on the BMR scheme performs favorably on the large benchmark dataset with 50 challenging videos. Abstract: In this paper, we present a simple yet effective Boolean map based representation that exploits connectivity cues for visual tracking. We describe a target object with histogram of oriented gradients and raw color features, of which each one is characterized by a set of Boolean maps generated by uniformly thresholding their values. The Boolean maps effectively encode multi-scale connectivity cues of the target with different granularities. The fine-grained Boolean maps capture spatially structural details that are effective for precise target localization while the coarse-grained ones encode global shape information that are robust to large target appearance variations. Finally, all the Boolean maps form together a robust representation that can be approximated by an explicit feature map of the intersection kernel, which is fed into a logistic regression classifier with online update, and the target location is estimated within a particle filter framework. The proposed representation scheme is computationally efficient and facilitates achieving favorable performance in terms of accuracy and robustness against the state-of-the-art tracking methods on the OTB50 and VOT2016 benchmark datasets.