[1]徐 康,龙 敏.增强尺度估计的特征压缩跟踪算法[J].红外技术,2018,40(12):1176-1181.[doi:10.11846/j.issn.1001_8891.201812010]
 XU Kang,LONG Min.Feature Compression Tracking Algorithm with Enhanced Scale Estimation[J].Infrared Technology,2018,40(12):1176-1181.[doi:10.11846/j.issn.1001_8891.201812010]
点击复制

增强尺度估计的特征压缩跟踪算法
分享到:

《红外技术》[ISSN:1001-8891/CN:CN 53-1053/TN]

卷:
40
期数:
2018年第12期
页码:
1176-1181
栏目:
出版日期:
2018-12-21

文章信息/Info

Title:
Feature Compression Tracking Algorithm with Enhanced Scale Estimation
文章编号:
1001-8891(20)12-1176-06
作者:
徐 康1龙 敏12
1. 长沙理工大学 计算机与通信工程学院,湖南 长沙 410114; 2. 长沙理工大学 综合交通运输大数据智能处理湖南省重点实验室,湖南 长沙 410114
Author(s):
XU Kang1LONG Min12
(1. School of Computer and Communication Engineering, Changsha University of Science and Technology, Changsha 410114, China; 2. Hunan Provincial Key Laboratory of Intelligent Processing of Big Data on Transportation, Changsha University of Science and Technology, Changsha 410114, China)
关键词:
相关滤波尺度估计特征压缩目标跟踪
Keywords:
correlation filteringscale estimationfeature compressionobject tracking
分类号:
TP391.41
DOI:
10.11846/j.issn.1001_8891.201812010
文献标志码:
A
摘要:
目标跟踪过程中,目标的尺度变大会引入更多的背景噪声,而在目标尺度变小时却采样不足,导致算法鲁棒性不强。为了实现复杂背景环境下可视目标的稳健跟踪,本文提出一种增强尺度估计的特征压缩跟踪算法,单独设置一个判别相关滤波器用于尺度估计,在线学习更新样本尺度,实时匹配最佳目标尺寸并更新特征采样块尺寸,对样本特征压缩降维并在线学习更新分类器参数,减小计算开销,提高跟踪稳健性。实验结果显示,算法可以适应目标的姿态及尺度变化,与已有类似算法相比,本文提出的算法具有更强的鲁棒性。
Abstract:
Additional background noise is introduced into a target tracking process as the scale of the target becomes larger. Furthermore, undersampling is achieved when the scale of the target becomes smaller, thereby resulting in an algorithm lacking robustness. To achieve robust tracking of visual targets in complex backgrounds, a feature compression tracking algorithm with enhanced scale estimation is proposed in this paper. A discriminant correlation filter is set up for scale estimation, and the sample size is updated online; therefore, the optimal target size is matched in real-time and the feature sample size is updated. In this algorithm, the sample feature is compressed and the classifier parameters are learned to reduce computation cost and improve tracking robustness. The experimental results show that the algorithm can adapt to the changes of pose and scale of the target. Compared with previous algorithms, this algorithm exhibits improved robustness.

参考文献/References:

[1]. FAN H, XIANG J. Robust visual tracking with multitask joint dictionary learning[J]. IEEE Transactions on Circuits & Systems for Video Technology, 2017, 27(5): 1018-1030.
[2]. ZHANG Tianzhu, Ghanem Bernard , LIU Si, et al. Robust visual tracking via structured multi-task sparse learning[J]. International Journal of Computer Vision, 2013, 101(2) : 367-383.
[3]. Choi A, Tavabi N, Darwiche A. Structured features in naive bayes classification[C]// Thirtieth Aaai Conference on Artificial Intelligence, 2016: 3233-3240.
[4]. HE Z, YI S, CHEUNG Y M, et al. Robust object tracking via key patch sparse representation[J]. IEEE Transactions on Cybernetics, 2017, 47(2): 354-364.
[5]. FAN H, XIANG J. Robust visual tracking with multitask joint dictionary learning[J]. IEEE Transactions on Circuits & Systems for Video Technology, 2017, 27(5): 1018-1030.
[6]. MEI X, LING H. Robust visual tracking using l(1) minimization[C] //IEEE Conference on Computer Vision of DBLP, 2009: 1436-1443.
[7]. Kalal Z, Mikolajczyk K, Matas J. Tracking-learning-detection[J]. IEEE Transactions on Pattern Analysis & Machine Intelligence, 2012, 34(7): 1409-1422.
[8]. ZHANG K, ZHANG L, YANG M H. Real-time compressive tracking[C]//European Conference on Computer Vision, Springer-Verlag, 2012: 864-877.
[9]. Bolme D S, Beveridge J R, Draper B A, et al. Visual object tracking using adaptive correlation filters[C]// Computer Vision and Pattern Recognition of IEEE, 2010: 2544-2550.
[10]. Henriques J F, Rui C, Martins P, et al. High-speed tracking with Kernelized correlation filters[J]. IEEE Transactions on Pattern Analysis & Machine Intelligence, 2015, 37(3): 583-596.
[11]. Arjoune Y, Kaabouch N, Ghazi H E, et al. Compressive sensing: performance comparison of sparse recovery algorithms[C]// Computing and Communication Workshop and Conference of IEEE, 2017: 1-7.
[12]. Candes E J, Tao T. Near-Optimal Signal Recovery From Random Projections: Universal Encoding Strategies?[M]. IEEE Press, 2006.
[13]. Baraniuk R, Davenport M, Devore R, et al. A simple proof of the restricted isometry property for random matrices[C]//Constr. Approx., 2008: 253-263.
[14]. Candes E J, Tao T. Decoding by linear programming[J]. IEEE Transactions on Information Theory, 2005, 51(12): 4203-4215.
[15]. Achlioptas D. Database-friendly random projections: Johnson-Linden- strauss with binary coins[J]. Journal of Computer & System Sciences, 2003, 66(4): 671-687.
[16]. ZHAO M, QIAN H M, RONG Y J, et al. Robust object tracking via sparse representation based on compressive collaborative Haar-like feature space[C/OL]// IEEE Conference of Audio, Language and Image Processing, https://www.researchgate.net/publication/ 262366524_ Ro- bust_ object_ tracking_ via_sparsity-based_ collaborative_model.
[17]. Tolstikhin I, Gelly S, Bousquet O, et al. Adagan: boosting generative models[C/OL]//31st Conference on Neural Information Processing Systems, https://arxiv.org/pdf/1701.02386.pdf.
[18]. ZHOU T, OUYANG Y, WANG R, et al. Particle filter based on real-time compressive tracking[C]//IEEE Conference on Audio, Language and Image Processing, 2017: 754-759.
[19]. ZHANG K, ZHANG L, YANG M H. Fast compressive fracking[J]. IEEE Transactions on Pattern Analysis & Machine Intelligence, 2014, 36(10): 2002-2015.
[20]. ZHANG K, ZHANG L, LIU Q, et al. Fast visual tracking via dense spatio-temporal context learning[C]//European Conference on Computer Vision, 2014: 127-141.

相似文献/References:

[1]华逸伦,石英,杨明东,等.基于背景抑制和前景抗干扰的多尺度跟踪算法[J].红外技术,2018,40(11):1098.[doi:10.11846/j.issn.1001_8891.201811014]
 HUA Yilun,SHI Ying,YANG Mingdong,et al.Multi-Scale Tracking Algorithm Based on Background Suppression and Foreground Anti-disturbance[J].Infrared Technology,2018,40(12):1098.[doi:10.11846/j.issn.1001_8891.201811014]

备注/Memo

备注/Memo:
收稿日期:2018-06-27;修订日期:2018-09-20.
作者简介:徐康(1992-),男,湖北省黄冈市人,硕士研究生,主要研究方向为计算机视觉与图像处理。E-mail:astonxk@qq.com。
通信作者:龙敏(1977-),女,湖南湘乡人,教授,博士,主要研究方向为混沌理论及应用、无线通信及安全。E-mail:80951404@qq.com。
基金项目:湖南省自然科学基金(15JJ2007)。
更新日期/Last Update: 2018-12-19