Bidirectional Tracking Scheme for Visual Object Tracking Based on Recursive Orthogonal Least Squares

Zhiyong Huang,Yuanlong Yu,Miaoxing Xu

doi:10.1109/access.2019.2951056

Zhiyong Huang, Yuanlong Yu + Show 1 more

Open Access

https://doi.org/10.1109/access.2019.2951056

Copy DOI

Abstract

Visual object tracking in unconstrained environments is a challenging task in computer vision. How to design an efficient discriminative feature representation is one challenging issue. To improve the adaptability of the tracker to large object appearance changes, the observation model needs to be updated online. However, a bad model update using inaccurate training samples can lead to model drift problem. Therefore, how to design an efficient online observation model and a model update strategy are two other challenging issues. This paper proposes the concatenation of histogram of oriented gradients variant (HOGv) and color histogram as the feature representation to balance discriminative power and efficiency. The single-hidden-layer feedforward neural network (SFNN) is used as an observation model, and the recursive orthogonal least squares (ROLS) algorithm is used to update the model online. A bidirectional tracking scheme is designed to alleviate the model drift problem during online tracking. The proposed bidirectional tracking scheme consists of three modules: the forward tracking module, the backward tracking module and the integration module. The forward tracking module first finds all the candidate regions, and then, the backward tracking module calculates the respective confidence of each candidate region according to historical information. Finally, the integration module integrates both of the first two modules' results to determine the final tracked object and the model update strategy for the current frame. Extensive evaluations of the existing tracking benchmarks have shown that the proposed tracking framework results in significant performance improvements compared with the base tracker, and it outperforms most of the state-of-the-art trackers.

Highlights

Visual object tracking, which is used to estimate the trajectory of a target specified in the initial frame, is a fundamental topic in computer vision [1], [2]
In visual object tracking, it is difficult to straightforwardly adopt convolutional neural networks (CNNs), since they require a large number of training samples, and there is only one labeled positive sample that is extracted from the initial frames
This paper proposes an efficient framework for visual object tracking

Summary

INTRODUCTION

Visual object tracking, which is used to estimate the trajectory of a target specified in the initial frame, is a fundamental topic in computer vision [1], [2]. Once all experts are polluted, the tracker will drift Compared with these methods [17]–[22], [29]–[34], the superiority of the proposed bidirectional tracking scheme is that the integration module of the scheme can determine the better model update strategy and the final tracked target for the current frame by fully considering both of the first two modules’ results and the spatial prior. This bidirectional tracking scheme is slightly similar to like ensemble learning techniques that combine the results that are achieved by weak trackers (i.e., the forward tracking module, the backward tracking module and the spatial prior) to produce a strong tracker (i.e., integration module) that is better than either of the weak trackers.

RELATE WORK

EXTRACTION OF HOGV AND COLOR HISTOGRAM DESCRIPTOR

ONLINE UPDATE PROCESS

EXPERIMENTS

EXPERIMENTAL SETUP

CONCLUSION