Robust visual tracker integrating adaptively foreground segmentation into multi-feature fusion framework
- PDF / 2,739,876 Bytes
- 24 Pages / 439.37 x 666.142 pts Page_size
- 98 Downloads / 167 Views
Robust visual tracker integrating adaptively foreground segmentation into multi-feature fusion framework Yi Zhang 1,2 & Guixi Liu 1,2
& Jiayu Gao
2,3
& Haoyang Zhang
1,2
Received: 13 July 2019 / Revised: 21 May 2020 / Accepted: 28 July 2020 # Springer Science+Business Media, LLC, part of Springer Nature 2020
Abstract
Existing Discriminative Correlation Filter (DCF) based methods suffer from the limitations of rectangular shape assumptions. Aiming at this issue, in this paper, we propose an effective tracking approach which integrates a pixel-wise foreground segmentation mask into the correlation filter within a multi-feature fusion framework. Specifically, we first propose a novel segmentation algorithm which combines the color histogram with the spatial prior. On this basis, we implement a target-masked correlation filter (TMCF) tracker by introducing the foreground mask into a ridge regression, which successfully suppresses unexpected background information inside the bounding box. Secondly, we apply the alternating direction method of multipliers (ADMM) to solve our TMCF model efficiently to obtain the closed-form solution. Finally, a complementary fusion tracker by the combining of TMCF and color histogram scores (fTMCFCH) is formulated, which is robust to deformations and illumination changes simultaneously. The fusion factor is determined adaptively by the reliability derived from the target resolution of the trackers separately in each frame. We perform extensive experiments on three benchmarks: OTB-2013, OTB-2015 and Temple-Color-128. The concrete experimental results demonstrate that our tracker outperforms several state-of-theart trackers. Keywords Foreground segmentation . Correlation filter tracking . Color histogram . Response maps fusion
* Guixi Liu [email protected]
1
School of Mechano-Electronic Engineering, Xidian University, Xi’an 710071, China
2
Shaanxi Key Laboratory of Integrated and Intelligent Navigation, Xi’an 710068, China
3
Xi’an Research Institute of Navigation Technology, Xi’an 710068, China
Multimedia Tools and Applications
1 Introduction Visual tracking is an important research topic with numerous applications including intelligent surveillance, human-machine interaction, autonomous navigation, etc. [40, 42]. The kernel purpose of generic tracking is to locate the target continuously in consecutive frames on the basis of the initial specified target. Despite the significant advances which have been made recently, tracking in complicated scenarios is still considered as a difficult task due to the diverse challenges from the background or the target itself such as shape deformation, illumination variation and background clutter [38]. Generally, tracking methods are divided into generative and discriminative tracking [41]. Generative methods describe tracking as a template matching process which is realized by looking for the candidate objects most similar to the target appearance within the target region. These theoretical techniques, such as sparse representation [28
Data Loading...
