基于模板更新的深层孪生网络跟踪算法

doi:10.3969/j.issn.1000-1565.2022.02.016

河北大学学报(自然科学版) ›› 2022, Vol. 42 ›› Issue (2): 217-224.DOI: 10.3969/j.issn.1000-1565.2022.02.016

• • 上一篇

基于模板更新的深层孪生网络跟踪算法

陈丽萍,苑侗侗,杨文柱,陈向阳,王思乐

收稿日期:2021-05-29 出版日期:2022-03-25 发布日期:2022-04-12
通讯作者: 杨文柱(1968—)
作者简介:陈丽萍(1974—),女,河北涞水人,河北大学讲师,主要从事机器视觉方面的研究.E-mail:mc_chenlp@sina.com
基金资助:
河北省自然科学基金青年项目(F2017201069)

A deep siamese network algorithm with template updating for object tracking

CHEN Liping, YUAN Tongtong, YANG Wenzhu, CHEN Xiangyang, WANG Sile

School of Cyber Security and Computer, Hebei University, Baoding 071002, China

Received:2021-05-29 Online:2022-03-25 Published:2022-04-12

摘要/Abstract

摘要： 针对跟踪过程中因尺度变化、遮挡及运动模糊等造成的目标定位不准确问题,在SiamFC(fully-convolutional siamese network)的跟踪框架基础上提出了一种具有高置信度模板更新机制的深层孪生网络目标跟踪算法.首先,主干网络采用ResNet-50残差网络进行特征提取,并融合多层特征图进行目标预测;其次,为避免模板频繁更新带来的模板漂移问题,构建了高置信度的模板更新模块.在OTB100数据集上的实验结果表明,相比基准算法,文中算法的跟踪成功率和精确度分别提高了3.4%和2.6%;在多种挑战因素下的对比实验表明,文中算法可以较好地抵抗目标遮挡、尺度变化、运动模糊等多种复杂因素带来的影响,有很好的鲁棒性.

关键词: 目标跟踪, 特征提取, 特征融合, 孪生网络, SiamFC

Abstract: To solve the problem of inaccurate target location caused by scale variation, occlusion and motion blur during the tracking process, a deep Siamese network target tracking algorithm with high confidence template updating mechanism is proposed based on the SiamFC(fully convolutional siamese network). First, the main network uses ResNet-50 residual network for feature extraction and multi-layer feature maps for target prediction; second, a high confidence template updateing module is constructed to avoid the template drift caused by frequent updating. Experimental results indicate that the success rate and tracking accuracy of the proposed algorithm are increased by about 3.4% and 2.6% respectively compared to the benchmark algorithm when running on the dataset of OTB100. The experiments under various challenging factors show that the proposed algorithm has good robustness, which can resist the effects of various complex factors effectively such as target occlusion, scale variation and motion blur.

Key words: object tracking, feature extraction, feature fusion, siamese network, SiamFC

中图分类号:

TP391

陈丽萍,苑侗侗,杨文柱,陈向阳,王思乐. 基于模板更新的深层孪生网络跟踪算法[J]. 河北大学学报(自然科学版), 2022, 42(2): 217-224.

CHEN Liping, YUAN Tongtong, YANG Wenzhu, CHEN Xiangyang, WANG Sile. A deep siamese network algorithm with template updating for object tracking[J]. Journal of Hebei University(Natural Science Edition), 2022, 42(2): 217-224.

参考文献

[1] 孟琭,杨旭.目标跟踪算法综述[J].自动化学报,2019, 45(7): 1244-126. DOI: 10.16383/j.aas.c180277.
[2] 陈云芳,吴懿,张伟.基于孪生网络结构的目标跟踪算法综述[J].计算机工程与应用,2020,56(6):10-18. DOI:10.3778/j.issn.1002-8331.1911-0127.
[3] BERTINETTO L, VALMADRE J, HENRIQUES J F, et al. Fully-convolutional siamese networks for object tracking[C] //Proceedings of the European Conference on Computer Vision(ECCV), 2016: 850-865. DOI:10.1007/978-3-319-48881-3_56.
[4] LI B, YAN J J, WU W, et al. High-performance visual tracking with siamese region proposal network[C] //Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition(CVPR), 2018: 8971-8980. DOI:10.1109/CVPR.2018.00935.
[5] FAN H, LING H. Siamese cascaded region proposal networks for Real-Time visual tracking[C] //Proceedings of the IEEE conference on Computer Vision and Pattern Recognition(CVPR), 2019: 7952-7961. DOI:10.1109/CVPR.2019.00814.
[6] ZHANG Z, PENG H. Deeper and wider siamese networks for real-time visual tracking[C] //Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition(CVPR), 2019: 4591-4600. DOI: 10.1109/CVPR.2019.00472.
[7] YU Y, XIONG Y, HUANG W, et al. Deformable siamese attention networks for visual object tracking[C] //Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition(CVPR), 2020:6728-6737. DOI:10.1109/CVPR42600.2020.00676.
[8] HE K, ZHANG X, REN S, et al. Deep residual learning for image recognition[C] //Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition(CVPR), 2016: 770-778. DOI:10.1109/CVPR.2016.90.
[9] WANG M, LIU Y, HUANG Z. Large margin object tracking with circulant feature maps[C] //Proceedings of the IEEE conference on Computer Vision and Pattern Recognition(CVPR), 2017: 4800-4808. DOI:10.1109/CVPR.2017.510.
[10] REAL E, SHLENS J, MAZZOCCHI S, et al. YouTube-bounding boxes:A large high-precision human-annotated dataset for object detection in video[C] //Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition(CVPR), 2017: 5296-5305. DOI: 10.1109/CVPR.2017.789.
[11] LIN T Y, MAIRE M, BELONGIE S, et al. Microsoft COCO: Common objects in context[C] //Proceedings of the European Conference on Computer Vision(ECCV), 2014: 740-755. DOI:10.1007/978-3-319-10602-1_48.
[12] RUSSAKOVSKY O, DENG J, SU H, et al. ImageNet large scale visual recognition challenge [J]. International Journal of Computer Vision, 2015, 115(3): 211-252. DOI:10.1007/s11263-015-0816-y.
[13] WU Y, LIM J, YANG M H. Object tracking benchmark [J]. IEEE Transactions on Pattern Analysis and Machine Intelligence, 2015, 37(9): 1834-1848. DOI:10.1109/TPAMI.2014.2388226.
[14] BERTINETTO L, VALMADRE J, GOLODETZ S, et al. Staple: Complementary learners for real-time tracking[C] //Proceedings of the IEEE conference on Computer Vision and Pattern Recognition(CVPR), 2016: 1401-1409. DOI:10.1109/CVPR.2016.156.
[15] MA C, YANG X K, ZHANG C Y, et al. Long-term correlation tracking[C] //Proceedings of the IEEE conference on Computer Vision and Pattern Recognition(CVPR), 2015: 5388-5396. DOI:10.1109/CVPR.2015.7299177.
[16] VALMADRE J, BERTINETTO L, HENRIQUES J F, et al. End-to-end representation learning for correlation filter based tracking[C] //Proceedings of the IEEE conference on Computer Vision and Pattern Recognition(CVPR), 2017: 2805-2813. DOI:10.1109/CVPR.2017.531.
[17] ZHANG J, MA S, SCLAROFF S. MEEM: Robust tracking via multiple experts using entropy minimization[C] //Proceedings of the European Conference on Computer Vision(ECCV), 2014: 188-203. DOI:10.1007/978-3-319-10599-4_13.
[18] DANELLJAN M, HÄGER G, KHAN F S, et al. Accurate scale estimation for robust visual tracking[C] //Proceeding of the British Machine Vision Conference(BMVC), 2014: 65.1-65.11. DOI:10.5244/C.28.65. (

基于模板更新的深层孪生网络跟踪算法

A deep siamese network algorithm with template updating for object tracking

PDF (PC)

可视化

摘要/Abstract

引用本文

使用本文

参考文献

相关文章 15

编辑推荐

Metrics

本文评价

[1]	刘振鹏,张庆文,李泽园,刘嘉航,董姝慧,赵永刚. 软件定义网络中基于贝叶斯ARTMAP的DDoS攻击检测模型[J]. 河北大学学报(自然科学版), 2021, 41(6): 728-733.
[2]	秦益文,封志宏,苏楠,马丁,李小菲. 基于深度置信网络的室内指纹定位算法[J]. 河北大学学报(自然科学版), 2021, 41(3): 321-328.
[3]	李凯,曹可凡,沈皓凝. 基于步态序列的跨视角步态识别[J]. 河北大学学报(自然科学版), 2021, 41(3): 311-320.
[4]	柴变芳,杨蕾,王建岭,李仁玲. 集成局部和全局关键特征的文本情感分类方法[J]. 河北大学学报(自然科学版), 2021, 41(2): 201-211.
[5]	张军,张敏,郝小可,解鹏. 基于多尺度特征融合的中分辨率遥感场景分类算法[J]. 河北大学学报(自然科学版), 2019, 39(6): 666-672.
[6]	李凯,曹可凡. 受限玻尔兹曼机的步态特征提取及其识别[J]. 河北大学学报(自然科学版), 2019, 39(6): 657-665.
[7]	王思乐,王铭羽,杨文柱,陈丽萍,陈向阳. 基于时空加权的多特征融合动作识别算法[J]. 河北大学学报(自然科学版), 2019, 39(1): 93-98.
[8]	石贵民,余文森,肖钟捷. 基于Gabor特征和协同神经网络的车牌识别方法[J]. 河北大学学报(自然科学版), 2016, 36(2): 210-217.
[9]	张玮,李晓苇,李光,张荣香. 二次特征提取法用于茶叶产地的识别[J]. 河北大学学报(自然科学版), 2014, 34(5): 479-484.
[10]	翟俊海,赵文秀,李塔. 一种简单有效的人脸识别方法[J]. 河北大学学报(自然科学版), 2013, 33(6): 654-660.
[11]	李凯,徐治平. 一种基于半监督学习的2DPCA人脸识别方法[J]. 河北大学学报(自然科学版), 2013, 33(4): 413-419.
[12]	翟俊海,翟梦尧,王华超. 基于小波变换和2DPCA的人脸识别[J]. 河北大学学报(自然科学版), 2010, 30(5): 574-579.
[13]	徐建民,刘清江,付婷婷,戴旭. 基于量化同义词关系的改进特征词提取方法[J]. 河北大学学报(自然科学版), 2010, 30(1): 97-101.
[14]	翟俊海,赵文秀,王熙照. 图像特征提取研究[J]. 河北大学学报(自然科学版), 2009, 29(1): 106-112.
[15]	翟俊海,张素芳,王熙照. 基于小波变换和支持向量机的图像分类[J]. 河北大学学报(自然科学版), 2007, 27(3): 317-321.