Issue |
JNWPU
Volume 37, Number 2, April 2019
|
|
---|---|---|
Page(s) | 323 - 329 | |
DOI | https://doi.org/10.1051/jnwpu/20193720323 | |
Published online | 05 August 2019 |
A New Video Tracking Algorithm Based on Multi-Complementary Features Fusion
基于多互补特征融合的视频跟踪算法
1
School of Electronics and Information, Northwestern Polytechnical University, Xi’an 710072, China
2
Science and Technology on Electro-Optic Control Laboratory, Luoyang 471000, China
Received:
18
April
2018
In order to make full use of the diversity of sample information in the tracking process and improve the generalization ability of the tracker, this paper integrates the object model prediction results on the basis of the Staple algorithm, and applies weighted bands to the simple linearity of different predictive response results in the algorithm. To the uncertainties, a new adaptive response factor graph fusion method with weight coefficients is proposed, which effectively improves the reliability of the video target tracking algorithm. Theoretical analysis and experimental simulation show that the proposed algorithm is more accurate and robust than the classical Staple algorithm, and it maintains high real-time performance.
摘要
为充分利用跟踪过程中样本信息的多样性,提高跟踪器的泛化能力,在Staple算法的基础上融入了基于轮廓特征的物体性模型预测结果,并针对该算法中对不同预测响应结果简单线性进行加权带来的不确定性,提出一种新的自适应权重系数的响应图融合方法,从而有效地提升了跟踪算法的可靠性。理论分析与实验仿真表明,所提算法在精准度和鲁棒性较经典的Staple算法有着较大提高,并且保持着较高的实时性。
Key words: correlation filters / object tracking / contour features / adaptive weights / feature fusion
关键字 : 相关滤波 / 目标跟踪 / 轮廓特征 / 自适应权重
© 2019 Journal of Northwestern Polytechnical University. All rights reserved.
This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/4.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.
Current usage metrics show cumulative count of Article Views (full-text article views including HTML views, PDF and ePub downloads, according to the available data) and Abstracts Views on Vision4Press platform.
Data correspond to usage on the plateform after 2015. The current usage metrics is available 48-96 hours after online publication and is updated daily on week days.
Initial download of the metrics may take a while.