首页 » 文章 » 文章详细信息
Mathematical Problems in Engineering Volume 2017 ,2017-10-17
Scale Adaptive Kernelized Correlation Filter Tracker with Feature Fusion
Research Article
Tongxue Zhou 1 , 2 , 3 Ming Zhu 1 Dongdong Zeng 1 , 2 , 3 Hang Yang 1
Show affiliations
DOI:10.1155/2017/1605959
Received 2017-05-13, accepted for publication 2017-07-06, Published 2017-07-06
PDF
摘要

Visual tracking is one of the most important components in numerous applications of computer vision. Although correlation filter based trackers gained popularity due to their efficiency, there is a need to improve the overall tracking capability. In this paper, a tracking algorithm based on the kernelized correlation filter (KCF) is proposed. First, fused features including HOG, color-naming, and HSV are employed to boost the tracking performance. Second, to tackle the fixed template size, a scale adaptive scheme is proposed which strengthens the tracking precision. Third, an adaptive learning rate and an occlusion detection mechanism are presented to update the target appearance model in presence of occlusion problem. Extensive evaluation on the OTB-2013 dataset demonstrates that the proposed tracker outperforms the state-of-the-art trackers significantly. The results show that our tracker gets a 14.79% improvement in success rate and a 7.43% improvement in precision rate compared to the original KCF tracker, and our tracker is robust to illumination variations, scale variations, occlusion, and other complex scenes.

授权许可

Copyright © 2017 Tongxue Zhou et al. 2017
This is an open access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

图表

Process of the feature fusion.

Response map of single feature and fused feature.

The scale adaptive scheme.

The original frames and response map of fz, respectively, (a) target without occlusion; (b) response map of fz without occlusion; (c) target with occlusion; (d) response map of fz in the presence of occlusion.

Success plots of videos with 8 different attributes for the top 8 trackers (the number in the title indicates the number of sequences).

Precision plots and success plots of OPE, TRE, and SRE for the top 8 trackers.

The results on three scale variation sequences.

The results on three illumination variation sequences.

The results on three occlusion sequences.

通讯作者

Tongxue Zhou.Chinese Academy of Sciences, Changchun Institute of Optics, Fine Mechanics and Physics, Changchun 130033, China, ciomp.ac.cn;The University of the Chinese Academy of Sciences, Beijing 100049, China, ucas.ac.cn;The Key Laboratory of Airborne Optical Imaging and Measurement, Chinese Academy of Sciences, Changchun 130033, China, cas.cn.zhoutongxue1992@163.com

推荐引用方式

Tongxue Zhou,Ming Zhu,Dongdong Zeng,Hang Yang. Scale Adaptive Kernelized Correlation Filter Tracker with Feature Fusion. Mathematical Problems in Engineering ,Vol.2017(2017)

您觉得这篇文章对您有帮助吗?
分享和收藏
30

是否收藏?

参考文献
[1] K. Briechle, U. D. Hanebeck. (2001). Template matching using fast normalized cross correlation. Aerospace/Defense Sensing, Simulation, and Controls:95-102. DOI: 10.1109/TPAMI.2013.230.
[2] Z. Kalal, J. Matas, K. Mikolajczyk. P-N learning: bootstrapping binary classifiers by structural constraints. :49-56. DOI: 10.1109/TPAMI.2013.230.
[3] D. Comaniciu, V. Ramesh, P. Meer. Real-time tracking of non-rigid objects using mean shift. :142-149. DOI: 10.1109/TPAMI.2013.230.
[4] G. An, H. Li, M. Li. (2015). Novel multi-window tracking based on model update. IEEE International Symposium on Computational Intelligence & Design:3-6. DOI: 10.1109/TPAMI.2013.230.
[5] J. F. Henriques, R. Caseiro, P. Martins, J. Batista. et al.(2015). High-speed tracking with kernelized correlation filters. IEEE Transactions on Pattern Analysis and Machine Intelligence.37(3):583-596. DOI: 10.1109/TPAMI.2013.230.
[6] X. Lan, Z. Shengping, P. C. Yuen. Robust joint discriminative feature learning for visual tracking. :3403-3410. DOI: 10.1109/TPAMI.2013.230.
[7] M. Danelljan, F. S. Khan, M. Felsberg, J. van de Weijer. et al.Adaptive color attributes for real-time visual tracking. :1090-1097. DOI: 10.1109/TPAMI.2013.230.
[8] T. B. Dinh, N. Vo, G. Medioni. Context tracker: exploring supporters and distracters in unconstrained environments. :1177-1184. DOI: 10.1109/TPAMI.2013.230.
[9] H. T. Nguyen, A. W. M. Smeulders. (2006). Robust tracking using foreground-background texture discrimination. International Journal of Computer Vision.69(3):277-293. DOI: 10.1109/TPAMI.2013.230.
[10] X. Lan, P. C. Yuen, R. Chellappa. Robust MIL-based feature template learning for object tracking. :4118-4125. DOI: 10.1109/TPAMI.2013.230.
[11] R. Liu, X. Lan, P. C. Yuen, G. C. Feng. et al.Robust visual tracking using dynamic feature weighting based on multiple dictionary learning. :2166-2170. DOI: 10.1109/TPAMI.2013.230.
[12] B. Babenko, S. Belongie, M. Yang. Visual tracking with online multiple instance learning. :983-990. DOI: 10.1109/TPAMI.2013.230.
[13] X. Lan, A. J. Ma, P. C. Yuen, R. Chellappa. et al.(2015). Joint sparse representation and robust feature-level fusion for multi-cue visual tracking. IEEE Transactions on Image Processing.24(12):5826-5841. DOI: 10.1109/TPAMI.2013.230.
[14] A. Smeulder, D. Chu, R. Cucchiara, S. Calderara. et al.(2014). Visual tracking: an experimental survey. IEEE Transactions on Pattern Analysis and Machine Intelligence.36(7):1442-1468. DOI: 10.1109/TPAMI.2013.230.
[15] F. Henriques, R. Caseiro, P. Martins, J. Batista. et al.Exploiting the circulant structure of tracking-by-detection with kernels. . DOI: 10.1109/TPAMI.2013.230.
[16] D. S. Bolme, J. R. Beveridge, B. A. Draper, Y. M. Lui. et al.Visual object tracking using adaptive correlation filters. :2544-2550. DOI: 10.1109/TPAMI.2013.230.
[17] A. Mahalanobis, B. V. K. V. Kuma, D. Casasent. (1987). Minimum average correlation energy filters. Applied Optics.26(17):3633-3640. DOI: 10.1109/TPAMI.2013.230.
[18] Y. Wu, J. Lim, M.-H. Yang. Online object tracking: a benchmark. :2411-2418. DOI: 10.1109/TPAMI.2013.230.
[19] Z.-Q. Hou, A.-Q. Huang, W.-S. Yu, X. Liu. et al.(2015). Visual object tracking method based on local patch model and model update. Journal of Electronics and Information Technology.37(6):1357-1364. DOI: 10.1109/TPAMI.2013.230.
[20] Y. Jiang, J. Ma. Combination features and models for human detection. :240-248. DOI: 10.1109/TPAMI.2013.230.
[21] B. Babenko, M. Yang, S. Belongie. Visual tracking with online multiple instance learning. :983-990. DOI: 10.1109/TPAMI.2013.230.
[22] S. Hare, A. Saffari, P. H. S. Torr. Struck: structured output tracking with kernels. :263-270. DOI: 10.1109/TPAMI.2013.230.
文献评价指标
浏览 217次
下载全文 23次
评分次数 0次
用户评分 0.0分
分享 30次