• 北大核心期刊(《中文核心期刊要目总览》2017版)
  • 中国科技核心期刊(中国科技论文统计源期刊)
  • JST 日本科学技术振兴机构数据库(日)收录期刊

留言板

尊敬的读者、作者、审稿人, 关于本刊的投稿、审稿、编辑和出版的任何问题, 您可以本页添加留言。我们将尽快给您答复。谢谢您的支持!

姓名
邮箱
手机号码
标题
留言内容
验证码

卷积视角下抗遮挡相关滤波跟踪方法

李丽惠 黄育明 丁灿 喻飞 陈颖频

李丽惠, 黄育明, 丁灿, 喻飞, 陈颖频. 卷积视角下抗遮挡相关滤波跟踪方法[J]. 微电子学与计算机, 2022, 39(8): 63-70. doi: 10.19304/J.ISSN1000-7180.2022.0081
引用本文: 李丽惠, 黄育明, 丁灿, 喻飞, 陈颖频. 卷积视角下抗遮挡相关滤波跟踪方法[J]. 微电子学与计算机, 2022, 39(8): 63-70. doi: 10.19304/J.ISSN1000-7180.2022.0081
LI Lihui, HUANG Yuming, DING Can, YU Fei, CHEN Yingpin. Anti-occlusion correlation filter tracking method from a convolution perspective[J]. Microelectronics & Computer, 2022, 39(8): 63-70. doi: 10.19304/J.ISSN1000-7180.2022.0081
Citation: LI Lihui, HUANG Yuming, DING Can, YU Fei, CHEN Yingpin. Anti-occlusion correlation filter tracking method from a convolution perspective[J]. Microelectronics & Computer, 2022, 39(8): 63-70. doi: 10.19304/J.ISSN1000-7180.2022.0081

卷积视角下抗遮挡相关滤波跟踪方法

doi: 10.19304/J.ISSN1000-7180.2022.0081
基金项目: 

闽南师范大学校长基金 KJ19019

闽南师范大学高级别项目 GJ19019

福建省自然科学基金项目 2020J05169

漳州职业技术学院校级课题 ZZY2021B129

大学生创新创业训练计划项目 202210402009

大学生创新创业训练计划项目 S202210402038

大学生创新创业训练计划项目 S202210402025

详细信息
    作者简介:

    李丽惠  女,(1982-),硕士,高级工程师.研究方向为图像处理、计算机视觉等

    通讯作者:

    陈颖频(通讯作者)   男,(1986-),博士,副教授.研究方向为图像处理、计算机视觉等.E-mail:110500617@163.com

  • 中图分类号: TP391.41

Anti-occlusion correlation filter tracking method from a convolution perspective

  • 摘要:

    针对当前相关滤波跟踪算法在目标进行旋转、快速运动和被遮挡时,易出现跟踪漂移甚至跟丢的问题,提出一种卷积视角下抗遮挡相关滤波跟踪方法.该方法在相关滤波算法框架的基础上,利用上下文感知方法增加背景信息,引入多模态历史池更新策略增强抗遮挡的跟踪性能.首先,设计出一套基于卷积视角的公式推导体系,巧妙地引入卷积定理在频域上求解滤波器,相比于现有文献中循环矩阵对角化的滤波器求解方法,该推导方法易于理解.然后,通过引入上下文相关信息,设计合理的能量泛函压制背景区域的响应值,达到更加稳健跟踪目标的目的.最后,建立历史多模态目标池,一旦相关响应最大的样本与历史模板池各多模态模板相似度低于人为设置阈值,则认定该帧出现遮挡,不进行模板池、外观模型、滤波器的更新,有效解决遮挡挑战下跟踪漂移的问题.将所提方法在OTB2015上进行测试,实验表明在目标旋转、快速运动、被遮挡等条件下,所提方法在保证准确跟踪的同时保持较高的速度,优于实验所提的其他方法.

     

  • 图 1  有无上下文感知模型对比实验

    Figure 1.  Comparison experiment with and without context perception model

    图 2  有无抗遮挡模块对比实验

    Figure 2.  Comparison experiment with or without anti-occlusion module

    图 3  OTB数据集不同序列的跟踪结果比较

    Figure 3.  Comparison of tracking results of different sequences in OTB dataset

    图 4  算法精确度和成功率综合比对图

    Figure 4.  A comprehensive comparison chart of the accuracy and success rate of the algorithm

    表  1  DCF_CA_AO算法表

    Table  1.   DCF_CA_AO Algorithm table

    算法表:DCF_CA_AO算法
    输入:第一帧训练样本x,历史多模态目标池T.
    输出:预测位置.
    1: For f=1:F do
    2: 根据式(16)训练相关滤波器${\hat{w}}_{l}^{*}$;
    3:    以第f-1帧为中心获取测试样本z
    4:    以式(11)检测样本z的响应;
    5:  以响应最大值确定第f帧的位置和最佳候选面片b
    6:    For n=1:N do
    7:        Simi(n)=coshtn, hb;
    8:    End
    9:        If max(Simi) < τ
    10:        x=x(old)
    11:    Else
    12:        以b的中心扩展获得当前帧训练样本$ {\tilde{x}}$;
    13:        x=(1-η)x(old)+η$ {\tilde{x}}$;
    14:        x(old)=x
    15:        将b更新到目标模板池T中;
    16:      End
    17:End
    下载: 导出CSV

    表  2  各跟踪算法在一些视频中的平均跟踪重叠率

    Table  2.   Average Tracking Overlap Rates for Each Tracking Algorithm in Some Videos

    视频序列 算法
    MTT SCM DSST STRUK KCF ASLA DCF_CA DCF_CA_AO
    Jogging-1 $\underset{\raise0.3em\hbox{$\smash{\scriptscriptstyle\thicksim}$}}{0.18}$ $\underset{\raise0.3em\hbox{$\smash{\scriptscriptstyle\thicksim}$}}{0.18}$ 0.19 0.17 0.19 $\underset{\raise0.3em\hbox{$\smash{\scriptscriptstyle\thicksim}$}}{0.18}$ $\underset{\raise0.3em\hbox{$\smash{\scriptscriptstyle\thicksim}$}}{0.18}$ 0.73
    lemming 0.29 0.14 0.33 0.48 $ \underset{\raise0.3em\hbox{$\smash{\scriptscriptstyle\thicksim}$}}{0.38}$ 0.14 0.23 0.54
    Human6 0.23 $ \underset{\raise0.3em\hbox{$\smash{\scriptscriptstyle\thicksim}$}}{0.34}$ 0.38 0.21 0.21 0.37 0.22 0.24
    Bird2 0.09 0.75 0.46 0.57 0.58 $ \underset{\raise0.3em\hbox{$\smash{\scriptscriptstyle\thicksim}$}}{0.60}$ $ \underset{\raise0.3em\hbox{$\smash{\scriptscriptstyle\thicksim}$}}{0.60}$ 0.65
    Panda 0.16 0.49 0.13 0.49 0.16 0.50 0.26 $ \underset{\raise0.3em\hbox{$\smash{\scriptscriptstyle\thicksim}$}}{0.32}$
    KiteSurf 0.30 0.27 $ \underset{\raise0.3em\hbox{$\smash{\scriptscriptstyle\thicksim}$}}{0.33}$ 0.64 0.47 0.25 0.25 0.32
    coke 0.44 0.33 $ \underset{\raise0.3em\hbox{$\smash{\scriptscriptstyle\thicksim}$}}{0.57}$ 0.61 0.55 0.17 $ \underset{\raise0.3em\hbox{$\smash{\scriptscriptstyle\thicksim}$}}{0.57}$ 0.62
    平均值 0.24 $ \underset{\raise0.3em\hbox{$\smash{\scriptscriptstyle\thicksim}$}}{0.36}$ 0.34 0.45 0.36 0.32 0.33 0.49
    下载: 导出CSV

    表  3  各跟踪算法在一些视频中的平均中心点误差

    Table  3.   Average center point error of each tracking algorithm in some videos

    视频序列 算法
    MTT SCM DSST STRUK KCF ASLA DCF_CA DCF_CA_AO
    Jogging-1 108.03 132.83 $ \underset{\raise0.3em\hbox{$\smash{\scriptscriptstyle\thicksim}$}}{87.90}$ 62.06 88.27 104.58 89.45 3.92
    lemming 165.43 185.72 81.89 37.75 $ \underset{\raise0.3em\hbox{$\smash{\scriptscriptstyle\thicksim}$}}{77.87}$ 178.82 150.03 64.03
    Human6 150.81 39.15 163.98 87.04 107.65 95.47 $ \underset{\raise0.3em\hbox{$\smash{\scriptscriptstyle\thicksim}$}}{86.37}$ 47.34
    Bird2 101.21 9.08 55.65 19.75 21.37 56.99 $ \underset{\raise0.3em\hbox{$\smash{\scriptscriptstyle\thicksim}$}}{18.77}$ 14.72
    Panda $ \underset{\raise0.3em\hbox{$\smash{\scriptscriptstyle\thicksim}$}}{20.16}$ 7.17 42.97 7.17 42.06 7.37 49.64 46.57
    KiteSurf 54.76 66.17 32.30 6.14 17.27 $ \underset{\raise0.3em\hbox{$\smash{\scriptscriptstyle\thicksim}$}}{25.86}$ 48.94 34.02
    coke 29.98 56.81 $ \underset{\raise0.3em\hbox{$\smash{\scriptscriptstyle\thicksim}$}}{12.79}$ 12.08 18.65 60.17 18.58 11.35
    平均值 90.05 70.99 68.21 33.14 $ \underset{\raise0.3em\hbox{$\smash{\scriptscriptstyle\thicksim}$}}{53.31}$ 75.61 65.97 31.71
    速度(fps) 0.41 0.36 20.40 20.2 155.72 4.86 75.30 $ \underset{\raise0.3em\hbox{$\smash{\scriptscriptstyle\thicksim}$}}{43.37}$
    下载: 导出CSV
  • [1] YU Q, DINH T B, MEDIONI G. Online tracking and reacquisition using co-trained generative and discriminative trackers[C]//10th European Conference on Computer Vision. Marseille: Springer, 2008: 678-691. DOI: 10.1007/978-3-540-88688-4_50.
    [2] LIU C J, WECHSLER H. Gabor feature based classification using the enhanced fisher linear discriminant model for face recognition[J]. IEEE Transactions on Image Processing, 2002, 11(4): 467-476. DOI: 10.1109/TIP.2002.999679.
    [3] BOLME D S, BEVERIDGE J R, DRAPER B A, et al. Visual object tracking using adaptive correlation filters[C]//Proceedings of the 2010 IEEE Computer Society Conference on Computer Vision and Pattern Recognition. San Francisco, CA, USA: IEEE, 2010: 2544-2550. DOI: 10.1109/CVPR.2010.5539960.
    [4] HENRIQUES J F, CASEIRO 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.
    [5] MUELLER M, SMITH N, GHANEM B. Context-aware correlation filter tracking[C]//Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition. Honolulu, HI, USA: IEEE, 2017: 1387-1395. DOI: 10.1109/CVPR.2017.152.
    [6] FAN J Q, SONG H H, ZHANG K H, et al. Real-time manifold regularized context-aware correlation tracking[J]. Frontiers of Computer Science, 2020, 14(2): 334-348. DOI: 10.1007/s11704-018-8104-y.
    [7] 陈法领, 丁庆海, 罗海波, 等. 采用时空上下文的抗遮挡实时目标跟踪[J]. 红外与激光工程, 2021, 50(1): 20200105. DOI: 10.3788/IRLA20200105.

    CHEN F L, DING Q H, LUO H B, et al. Anti-occlusion real time target tracking algorithm employing spatio-temporal context[J]. Infrared and Laser Engineering, 2021, 50(1): 20200105. DOI: 10.3788/IRLA20200105.
    [8] 虞虹玲, 陈颖频, 许艳萍, 等. 局部模板更新逆向联合稀疏表示目标跟踪算法[J]. 图学学报, 2022, 43(1): 60-69. DOI: 10.11996/JG.j.2095-302X.2022010060.

    YU H L, CHEN Y P, XU Y P, et al. Object tracking of reverse joint sparse representation with local template update[J]. Journal of Graphics, 2022, 43(1): 60-69. DOI: 10.11996/JG.j.2095-302X.2022010060.
    [9] 陈颖频, 程祝媛, 王灵芝, 等. 基于动态坐标和图解法的卷积积分计算[J]. 闽南师范大学学报(自然科学版), 2016, 29(1): 30-38. DOI: 10.16007/j.cnki.issn2095-7122.2016.01.006.

    CHEN Y P, CHENG Z Y, WANG L Z, et al. The computation of convolution based on the dynamic coordinate and graphical method[J]. Journal of Minnan Normal University (Natural Science), 2016, 29(1): 30-38. DOI: 10.16007/j.cnki.issn2095-7122.2016.01.006.
    [10] SHI X C, LING H B, PANG Y, et al. Rank-1 tensor approximation for high-order association in multi-target tracking[J]. International Journal of Computer Vision, 2019, 127(8): 1063-1083. DOI: 10.1007/s11263-018-01147-z.
    [11] 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.
    [12] 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. Las Vegas, NV, USA: IEEE, 2016: 1401-1409. DOI: 10.1109/CVPR.2016.156.
    [13] JIA X, LU H C, YANG M H. Visual tracking via adaptive structural local sparse appearance model[C]//Proceedings of the IEEE Conference on computer Vision & Pattern Recognition. Providence: IEEE, 2012: 1822-1829. DOI: 10.1109/CVPR.2012.6247880.
    [14] HARE S, GOLODETZ S, SAFFARI A, et al. Struck: structured output tracking with kernels[J]. IEEE Transactions on Pattern Analysis and Machine Intelligence, 2016, 38(10): 2096-2109. DOI: 10.1109/TPAMI.2015.2509974.
    [15] HENRIQUES J F, CASEIRO R, MARTINS P, et al. Exploiting the circulant structure of tracking-by-detection with kernels[C]//12th European Conference on Computer Vision. Florence: Springer, 2012: 702-715. DOI: 10.1007/978-3-642-33765-9_50.
    [16] ZHANG K H, ZHANG L, YANG M H. Real-time compressive tracking[C]//Proceedings of the 12th European Conference on Computer Vision. Florence: Springer, 2012: 864-877. DOI: 10.1007/978-3-642-33712-3_62.
    [17] DANELLJAN M, HÄGER G, KHAN F S, et al. Discriminative scale space tracking[J]. IEEE Transactions on Pattern Analysis and Machine Intelligence, 2017, 39(8): 1561-1575. DOI: 10.1109/tpami.2016.2609928.
    [18] ZHONG W, LU H C, YANG M H. Robust object tracking via sparsity-based collaborative model[C]//Proceedings of the 2012 IEEE Conference on Computer Vision and Pattern Recognition. Providence, Rhode Island, USA: IEEE, 2012: 1838-1845. DOI: 10.1109/cvpr.2012.6247882.
    [19] ZHANG T Z, GHANEM B, LIU S, et al. Robust visual tracking via multi-task sparse learning[C]//IEEE Computer Vision and Pattern Recognition (CVPR). Providence: IEEE, 2012: 2042-2049. DOI: 10.1109/CVPR.2012.6247908.
  • 加载中
图(4) / 表(3)
计量
  • 文章访问数:  148
  • HTML全文浏览量:  45
  • PDF下载量:  6
  • 被引次数: 0
出版历程
  • 收稿日期:  2022-01-27
  • 修回日期:  2022-03-06
  • 网络出版日期:  2022-08-15

目录

    /

    返回文章
    返回