基于核相关滤波与特征融合的分块跟踪算法

doi:10.16088/j.issn.1001-6600.2020.05.002

广西师范大学学报（自然科学版） ›› 2020, Vol. 38 ›› Issue (5): 12-23.doi: 10.16088/j.issn.1001-6600.2020.05.002

基于核相关滤波与特征融合的分块跟踪算法

张灿龙^1,2*, 李燕茹¹, 李志欣^1,2, 王智文³

1.广西师范大学广西多源信息挖掘与安全重点实验室, 广西桂林541004;
2.广西区域多源信息集成与智能信息处理协同创新中心, 广西桂林541004;
3.广西科技大学计算机科学与通信工程学院, 广西柳州545006

收稿日期:2020-01-20 出版日期:2020-09-25 发布日期:2020-10-09
通讯作者: 张灿龙(1975—), 男, 湖南双峰人, 广西师范大学教授, 博导。E-mail:zcltyp@163.com
基金资助:
国家自然科学基金(61866004,61966004,61962007); 广西自然科学基金(2018GXNSFDA281009, 2017GXNSFAA198365, 2019GXNSFDA245018, 2018GXNSFDA294001); 广西“八桂学者”创新研究团队; 广西多源信息挖掘与安全重点实验室基金(20-A-03-01); 广西师范大学计算机科学与信息工程学院创新项目(JXYJSKT-2019-002)

Block Target Tracking Based on Kernel Correlation Filter and Feature Fusion

ZHANG Canlong^1,2*, LI Yanru¹, LI Zhixin^1,2, WANG Zhiwen³

1. Guangxi Key Lab of Multi-source Information Mining and Security, Guangxi Normal University, Guilin Guangxi 541004, China;
2. Guangxi Collaborative Innovation Center of Multi-source Information Integration and Intelligent Processing, Guilin Guangxi 541004, China;
3. College of Computer Science and Communication Engineering, Guangxi University of Science and Technology, Liuzhou Guangxi 545006, China

Received:2020-01-20 Online:2020-09-25 Published:2020-10-09

摘要/Abstract

摘要： 为了使核相关滤波跟踪器能处理目标遮挡和形变问题,本文提出一种基于核相关滤波与多特征融合的自适应分块跟踪算法。首先,根据目标的大小及长宽比对其进行自适应分块,并对各子块提取方向梯度直方图和颜色特征;然后,利用融合后的特征对每个子块进行核相关滤波跟踪,得到各子块的最大响应位置;接着,通过子块与原目标间的几何关系得到其对应的候选目标位置;最后,通过对各候选目标位置加权平均得到最终的目标响应位置。此外,以子块响应图的峰值旁瓣比和其响应位置与目标最终位置的欧氏距离为判据来判断该子块的有效性,对于有效的子块进行自适应更新。在多个图像序列的实验结果表明,本文方法能较好地处理目标遮挡等问题,且其性能总体上要优于被比较的方法。

关键词: 目标跟踪, 相关滤波, 目标分块, 特征融合, 模型更新

Abstract: In order to make the tracker based kernel correlation filter to overcome occlusion and deformation, an adaptive block tracking algorithm is proposed by using kernel correlation filtering and multi-feature fusion. Firstly, adaptively divide it into blocks based on the size and height-width ratios of the target and extract the Histogram of oriented gradient (HOG) and color name (CN) features for each block; Secondly, represent the block with the fusion of HOG and CN, and use the kernel correlation filtering tracker to get the position with maximum response of each block; Thirdly, calculate the coordinate of candidate target through the coordinate geometric relationship between each block and the original target; Moreover, the final target position is obtained by weighted average of the coordinates of all coordinates. Finally, the peak to side lobe ratio(PSR) of a blocks response curve as well as the distance between its response position and the final target position are used to judge its validity, and the valid blocks are adaptively updated. The experimental results on several data set show that the proposed method performs well on occlusion and deformation, and its comprehensive performance is better than other methods.

Key words: target tracking, correlation filtering, target block, feature fusion, model update

中图分类号:

TP391.41

张灿龙, 李燕茹, 李志欣, 王智文. 基于核相关滤波与特征融合的分块跟踪算法[J]. 广西师范大学学报（自然科学版）, 2020, 38(5): 12-23.

ZHANG Canlong, LI Yanru, LI Zhixin, WANG Zhiwen. Block Target Tracking Based on Kernel Correlation Filter and Feature Fusion[J]. Journal of Guangxi Normal University(Natural Science Edition), 2020, 38(5): 12-23.

参考文献

[1] OJHA S, SAKHARE S. Image processing techniques for object tracking in video surveillance:a survey[C] // 2015 International Conference on Pervasive Computing (ICPC). Piscataway, NJ:IEEE Press, 2015:302-309. DOI: 10.1109/PERVASIVE.2015.7087180.
[2] 黄凯奇, 陈晓棠, 康运锋, 等. 智能视频监控技术综述[J]. 计算机学报, 2015, 38(6): 1093-1118. DOI: 10.11897/SP.J.1016.2015.01093.
[3] MUELLER M, SMITH N, GHANEM B. A benchmark and simulator for UAV tracking[C] // Computer Vision-ECCV 2016: 14th European Conference, Amsterdam, The Netherlands, October 11-14, 2016, Proceedings, Part I.Berlin: Springer, 2016: 445-461. DOI: 10.1007/978-3-319-46448-0_27.
[4] BOULAHIA S Y, ANQUETIL E, MULTON F, et al. Dynamic hand gesture recognition based on 3D pattern assembled trajectories[C] // 2017 Seventh International Conference on Image Processing Theory, Tools and Applications. Piscataway, NJ: IEEE Press, 2017: 1-6. DOI: 10.1109/IPTA.2017.8310146.
[5] SHIEH W Y, HSU C C J, WANG T H. Vehicle positioning and trajectory tracking by infrared signal-direction discrimination for short-range vehicle-to-infrastructure communication systems[J]. IEEE Transactions on Intelligent Transportation Systems, 2018, 19(2): 368-379. DOI: 10.1109/TITS.2017.2697041.
[6] CAI Y X, LI L, NI S L, et al. Moving vehicle detection based on dense SIFT and extreme learning machine for visual surveillance[C] // 2015 IEEE International Conference on Robotics and Biomimetics. Piscataway, NJ: IEEE Press, 2015: 1614-1618. DOI: 10.1109/ROBIO.2015.7419002.
[7] CLAUS P, OMAR A M S, PEDRIZZETTI G, et al. Tissue tracking technology for assessing cardiac mechanics: principles, normal values, and clinical applications[J]. JACC: Cardiovascular Imaging, 2015, 8(12): 1444-1460. DOI: 10.1016/j.jcmg.2015.11.001.
[8] UNGI T, ABOLMAESUMI P, JALAL R, et al. Spinal needle navigation by tracked ultrasound snapshots[J]. IEEE Transactions on Biomedical Engineering, 2012, 59(10): 2766-2772. DOI: 10.1109/TBME.2012.2209881.
[9] LI H X, SHEN C H, SHI Q F. Real-time visual tracking using compressive sensing[C] // Proceedings of the 2011 IEEE Conference on Computer Vision and Pattern Recognition. 2011: 1305-1312.DOI: 10.1109/CVPR.2011.5995483.
[10] 卢湖川, 李佩霞, 王栋. 目标跟踪算法综述[J]. 模式识别与人工智能, 2018, 31(1): 61-76.DOI: 10.16451/j.cnki. issn1003-6059.201801006.
[11] 张灿龙, 唐艳平, 李志欣, 等. 基于二阶空间直方图的双核跟踪[J]. 电子与信息学报, 2015, 37(7): 1660-1666. DOI: 10.11999/JEIT141321.
[12] 黄宏图, 毕笃彦, 查宇飞, 等. 基于笛卡尔乘积字典的稀疏编码跟踪算法[J]. 电子与信息学报, 2015, 37(3): 516-521. DOI: 10.11999/JEIT140931.
[13] ZHOU T R, OUYANG Y N, WANG R, et al. Particle filter based on real-time compressive tracking[C] // 2016 International Conference on Audio, Language and Image Processing. Piscataway, NJ: IEEE Press, 2016: 754-759. DOI: 10.1109/ICALIP.2016.7846666.
[14] YUN X, JING Z L, XIAO G, et al. A compressive tracking based on time-space Kalman fusion model[J]. Science China: Information Sciences, 2016, 59(1): 012106. DOI: 10.1007/s11432-015-5356-0.
[15] WEN C B, CAI Y Z, LIU Y R, et al. A reduced-order approach to filtering for systems with linear equality constraints[J]. Neurocomputing, 2016, 193: 219-226. DOI: 10.1016/j.neucom.2016.02.020.
[16] IMANI M, BRAGA-NETO U M. Particle filters for partially-observed boolean dynamical systems[J]. Automatica, 2018, 87: 238-250. DOI: 10.1016/j.automatica.2017.10.009.
[17] ALI A, JALIL A, NIU J W, et al. Visual object tracking: classical and contemporary approaches[J]. Frontiers of Computer Science, 2016, 10(1): 167-188.DOI: 10.1007/s11704-015-4246-3.
[18] BOLME D S, BEVERIDGE J R, DRAPER B A, et al. Visual object tracking using adaptive correlation filters[C] // 2010 IEEE Computer Society Conference on Computer Vision and Pattern Recognition.Los Alamtios, CA: IEEE Computer Society, 2010: 2544-2550. DOI: 10.1109/CVPR.2010.5539960.
[19] HENRIQUES J F, CASEIOR R, MARTINS P, et al. Exploiting the circulant structure of tracking-by-detection with kernels[C] // Computer Vision-ECCV 2012: 12th European Conference on Computer Vision, Florence, Italy, October 7-13, 2012, Proceedings, Part IV. Berlin: Springer, 2012: 702-715. DOI: 10.1007/978-3-642-33765-9_50.
[20] HENRIQUES J F, CASEIOR R, MARTINS P, et al. High-speed tracking with kernelized correlation filters[J]. IEEE Transactions on Pattern Analysis and Machine Intelligence, 2015, 37(3): 583-596. DOI: 10.1109/TPAMI.2014.2345390.
[21] DANELLJAN M, KHAN F S, FELSBERG M, et al. Adaptive color attributes for real-time visual tracking[C] // 2014 IEEE Conference on Computer Vision and Pattern Recognition. Los Alamtios, CA: IEEE Computer Society, 2014: 1090-1097. DOI: 10.1109/CVPR.2014.143.
[22] DANELLJAN M,HÄGER G, KHAN F S, et al. Accurate scale estimation for robust visual tracking[C] // Proceedings of the British Machine Vision Conference. Surrey: BMVC Press, 2014: 1-11. DOI: 10.5244/C.28.65.
[23] LI Y, ZHU J K, HOI S C H. Reliable patch trackers: robust visual tracking by exploiting reliable patches[C] // 2015 IEEE Conference on Computer Vision and Pattern Recognition. Los Alamtios, CA: IEEE Computer Society, 2015: 353-361. DOI: 10.1109/CVPR.2015.7298632.
[24] GÖNEN M, ALPAYDIN E. Multiple kernel learning algorithms[J]. Journal of Machine Learning Research, 2011, 12: 2211-2268.
[25] FELZENSAWALB P F, GIRSHICK R B,McALLESTER D, et al. Object detection with discriminatively trained part-based models[J]. IEEE Transactions on Pattern Analysis and Machine Intelligence, 2010, 32(9): 1627-1645. DOI: 10.1109/TPAMI.2009.167.
[26] VAN DE WEIJER J, SCHMID C, VERBEEK J, et al. Learning color names for real-world applications[J]. IEEE Transactions on Image Processing, 2009, 18(7): 1512-1523. DOI: 10.1109/TIP.2009.2019809.
[27] WU Y, LIM J W, YANG M H. Online object tracking: a benchmark[C] // 2013 IEEE Conference on Computer Vision and Pattern Recognition. Los Alamtios, CA: IEEE Computer Society, 2013: 2411-2418. DOI: 10.1109/CVPR.2013.312.

Metrics

Viewed

Full text

Abstract

Cited

Shared

Discussed

基于核相关滤波与特征融合的分块跟踪算法

Block Target Tracking Based on Kernel Correlation Filter and Feature Fusion

PDF (PC)

摘要/Abstract

引用本文

使用本文

参考文献

相关文章 2

Metrics

本文评价

推荐阅读 10

[1]	白捷, 高海力, 王永众, 杨来邦, 项晓航, 楼雄伟. 基于多路特征融合的Faster R-CNN与迁移学习的学生课堂行为检测[J]. 广西师范大学学报（自然科学版）, 2020, 38(5): 1-11.
[2]	马先兵, 孙水发, 覃音诗, 郭青, 夏平. 基于粒子滤波的on-line boosting目标跟踪算法[J]. 广西师范大学学报（自然科学版）, 2013, 31(3): 100-105.