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

融合判别性与细粒度特征的抗遮挡红外目标跟踪算法

吴捷 马小虎

吴捷, 马小虎. 融合判别性与细粒度特征的抗遮挡红外目标跟踪算法[J]. 红外技术, 2022, 44(11): 1139-1145.
引用本文: 吴捷, 马小虎. 融合判别性与细粒度特征的抗遮挡红外目标跟踪算法[J]. 红外技术, 2022, 44(11): 1139-1145.
WU Jie, MA Xiaohu. Anti-Occlusion Infrared Target Tracking AlgorithmBased on Fusion of Discriminant and Fine-Grained Features[J]. Infrared Technology , 2022, 44(11): 1139-1145.
Citation: WU Jie, MA Xiaohu. Anti-Occlusion Infrared Target Tracking AlgorithmBased on Fusion of Discriminant and Fine-Grained Features[J]. Infrared Technology , 2022, 44(11): 1139-1145.

融合判别性与细粒度特征的抗遮挡红外目标跟踪算法

基金项目: 

国家自然科学基金 61402310

江苏省自然科学基金 BK20141195

泰州职业技术学院重点科研项目 1821819039

详细信息
    作者简介:

    吴捷(1982-),男,副教授,主要研究方向:视觉目标跟踪。E-mail: 37323736@qq.com

  • 中图分类号: TN911.73

Anti-Occlusion Infrared Target Tracking AlgorithmBased on Fusion of Discriminant and Fine-Grained Features

  • 摘要: 针对现有热红外目标跟踪算法难以处理相似物干扰和目标遮挡的问题,引入MMNet(Multi-task Matching Network)算法中的多任务框架获取热红外目标特定的判别性特征和细粒度特征,并将这两种特征相互融合,用于在类间和类内识别热红外对象。此外,利用峰值旁瓣比动态设置模型更新参数以更高效地获取目标变化信息并对跟踪结果进行评估。对于不可靠跟踪结果利用卡尔曼滤波对目标位置进行预测。在LSOTB-TIR(Large-Scale Thermal Infrared Object Tracking Benchmark)红外数据集上的实验结果表明,提出的改进算法性能较好,相比MMNet跟踪精确度和成功率分别提高了5.7%和4.2%,且能有效应对遮挡、变形等挑战,可以应用于红外目标跟踪领域。
  • 图  1  多任务匹配网络(MMNet)框架图

    Figure  1.  Architecture of Multi-task Matching Network (MMNet)

    图  2  细粒度感知网络(FANet)的体系结构

    Figure  2.  Architecture of Fine-grained Aware Network (FANet)

    图  3  8种算法在LSOTB-TIR的距离精度曲线图和成功率曲线

    Figure  3.  Distance accuracy curves and success rate curves of eight algorithms in LSOTB-TIR

    图  4  8种算法在LSOTB-TIR的6种挑战性场景下精确度曲线图

    Figure  4.  Accuracy curve of eight algorithms in six challenging scenarios of LSOTB-TIR

    图  5  本文算法与MMNet算法跟踪结果比较

    Figure  5.  Comparison of tracking results between our algorithm and MMNet

  • [1] 张晋, 王元余, 林丹丹, 等. 基于相关滤波的红外目标跟踪抗遮挡处理[J]. 红外技术, 2022, 44(3): 277-285. http://hwjs.nvir.cn/article/id/98939f6c-0de2-4692-9c34-9eabbb68205e

    ZHANG Jin, WANG Yuanyu, LIN Dandan, et al. Anti-occlusion process of infrared target tracking based on correlation filters[J]. Infrared Technology, 2022, 44(3): 277-285. http://hwjs.nvir.cn/article/id/98939f6c-0de2-4692-9c34-9eabbb68205e
    [2] Gundogdu E, Koc A, Solmaz B, et al. Evaluation of feature channels for correlation-filter-based visual object tracking in infrared spectrum[C]// 2016 IEEE Conference on Computer Vision and Pattern Recognition Workshops(CVPRW), 2016: 290-298.
    [3] LIU Q, LU X H, HE Z Y, et al. Deep convolutional neural networks for thermal infrared object tracking[J]. Knowledge-Based Systems, 2017, 134: 189-198. doi:  10.1016/j.knosys.2017.07.032
    [4] GAO P, MA Y, SONG K, et al. Large margin structured convolution operator for thermal infrared object tracking[C]// IEEE International Conference on Pattern Recognition, 2018: 2380-2385.
    [5] LI X, LIU Q, FAN Nana, et al. Hierarchical spatial-aware Siamese network for thermal infrared object tracking[J]. Knowledge-Based Systems, 2019, 166: 71-81. doi:  10.1016/j.knosys.2018.12.011
    [6] LIU Q, LI X, HE Z Y, et al. Learning deep multi-level similarity for thermal infrared object tracking[J]. IEEE Transaction on Multimedia, 2021, 23: 2124-2126.
    [7] 李畅, 杨德东, 宋鹏, 等. 基于全局感知孪生网络的红外目标跟踪[J]. 光学学报, 2021, 41(6): 0615002-1-0615002-11. https://www.cnki.com.cn/Article/CJFDTOTAL-GXXB202106019.htm

    LI Chang, YANG Dedong, SONG Pen, et al. Global-Aware siamese network for thermal infrared object tracking[J]. Acta Optica Sinica, 2021, 41(6): 0615002-1-0615002-11. https://www.cnki.com.cn/Article/CJFDTOTAL-GXXB202106019.htm
    [8] LIU Q, LI X, HE Z Y, et al. Multi-task driven feature models for thermal infrared tracking[C]//Proceedings of the 34th AAAI Conference on Artificial Intelligence, 2020: 11604-11611.
    [9] Valmadre J, Bertinetto L, Henriques J, et al. End-to-end representation learning for correlation filter based tracking[C]//2017 IEEE Conference on Computer Vision and Pattern Recognition (CVPR), 2017: 5000-5008.
    [10] 王鹏, 孙梦宇, 王海燕, 等. 一种目标响应自适应的通道可靠性跟踪算法[J]. 电子与信息学报, 2020, 42(8): 1950-1958. https://www.cnki.com.cn/Article/CJFDTOTAL-DZYX202008018.htm

    WANG Peng, SUN Mengyu, WANG Haiyan. An object tracking algorithm with channel reliability and target response adaptation[J]. Journal of Electronics & Information Technology, 2020, 42(8): 1950-1958. https://www.cnki.com.cn/Article/CJFDTOTAL-DZYX202008018.htm
    [11] 刘耀胜, 廖育荣, 林存宝. 基于核相关滤波的视频卫星目标跟踪算法[J]. 火力与指挥控制, 2022, 47(2): 49-55. https://www.cnki.com.cn/Article/CJFDTOTAL-HLYZ202202009.htm

    LIU Yaosheng, LIAO Yurong, LIN Cunbao. Video satellite object tracking algorithm based on kernel correlation filter[J]. Fire Control & Command Control, 2022, 47(2): 49-55. https://www.cnki.com.cn/Article/CJFDTOTAL-HLYZ202202009.htm
    [12] MA C, HUANG J-B, YANG X, et al. Hierarchical convolutional features for visual racking[C]//IEEE International Conference on Computer Vision, 2015: 3074-3082.
    [13] WANG N, ZHOU W, TIAN Q, et al. Multi-cue correlation filters for robust visual tracking[C]// IEEE Conference on Computer Vision and Pattern Recognition, 2018: 4844-4853.
    [14] Danelljan M, Hager G, Shahbaz Khan F, et al. Learning spatially regularized correlation filters for visual tracking[C]//IEEE International Conference on Computer Vision, 2015: 4310-4318.
    [15] Russakovsky O, Deng J, Su H, et al. ImageNet large scale visual recognition challenge[J]. International Journal Computer Vision, 2015, 115(3): 211-252. doi:  10.1007/s11263-015-0816-y
    [16] LIU Q, LI X, LI C L. LSOTB-TIR: A large-scale high-diversity thermal infrared object tracking benchmark[C/OL]//Proceedings of the 28th ACM International Conference on Multimedia, 2020, https://arxiv.org/abs/2008.00836.
  • 加载中
图(5)
计量
  • 文章访问数:  121
  • HTML全文浏览量:  95
  • PDF下载量:  33
  • 被引次数: 0
出版历程
  • 收稿日期:  2022-06-18
  • 修回日期:  2022-08-31
  • 刊出日期:  2022-11-20

目录

    /

    返回文章
    返回