首页 > 论文 > 激光与光电子学进展 > 56卷 > 23期(pp:231503--1)

基于时间正则化及背景感知的滤波器跟踪

Filter Tracking Based on Time Regularization and Background-Aware

  • 摘要
  • 论文信息
  • 参考文献
  • 被引情况
  • PDF全文
分享:

摘要

针对相关滤波器(CF)的目标背景因没有根据时间建模而导致的性能不佳的问题,在方向梯度直方图(HOG)的基础上,提出一种基于时间正则化及背景感知的滤波器跟踪算法。从真实的背景中提取训练样本,通过增加训练样本来增强滤波器的分类能力;引入时间正则化,构建遮挡情况下目标重定位模块;采用交替方向乘子法(ADMM)优化求解目标,降低计算复杂度;采用线性插值策略来更新目标的位置和尺度。采用目标跟踪基准(OTB-2015)数据集中的100个视频序列与评价标准对本文所提出的算法进行性能测试。实验结果表明,基于时间正则化及背景感知的滤波器跟踪算法的精确度得分达到0.801,成功率得分为0.762,相比核相关滤波器(KCF)算法分别提高了20%和46.8%。本文算法能很好解决目标发生平面外旋转、目标被遮挡、背景嘈杂等情况下的视觉跟踪问题,具有良好的应用前景和较大的使用价值。

Abstract

This study proposes a filter tracking algorithm based on the direction gradient histogram using time regularization and background-aware to overcome the problem of target background of the correlation filter (CF) having no optimal performance without time modeling. The training samples are firstly extracted from the real background, and classification ability of the filter is enhanced by adding the training samples. Subsequently, time regularization is introduced to construct the target relocation module under occlusion. In addition, the alternating direction multiplier method is used to optimize the solution target and reduce the computational complexity. Finally, a linear interpolation strategy is used to update the target location and scale. The proposed algorithm uses 100 video sequences and evaluation criteria in object tracking benchmark (OTB-2015) dataset for performance testing. Experimental results show that the accuracy score of filter tracking algorithm using time regularization and background-aware reaches 0.801 and success rate score is 0.762, which are 20% and 46.8% higher, respectively, compared to those of the kernelized correlation filter (KCF) algorithm. The proposed algorithm can solve the visual-tracking problem of off-plane rotation, occlusion, and background ambiguity, which has wide application prospects and use value.

Newport宣传-MKS新实验室计划
补充资料

DOI:10.3788/LOP56.231503

所属栏目:机器视觉

收稿日期:2019-04-26

修改稿日期:2019-06-03

网络出版日期:2019-12-01

作者单位    点击查看

刘明明:西北师范大学物理与电子工程学院, 甘肃 兰州 730030
裴东:西北师范大学物理与电子工程学院, 甘肃 兰州 730030甘肃省智能信息技术与应用工程研究中心, 甘肃 兰州 730030
刘举:西北师范大学物理与电子工程学院, 甘肃 兰州 730030
祝东辉:西北师范大学物理与电子工程学院, 甘肃 兰州 730030
孙浩翔:西北师范大学物理与电子工程学院, 甘肃 兰州 730030

联系人作者:刘明明(651766323@qq.com); 裴东(peidong@nwnu.edu.cn);

【1】Wu Y, Lim J, Yang M H. Object tracking benchmark [J]. IEEE Transactions on Pattern Analysis and Machine Intelligence. 2015, 37(9): 1834-1848.

【2】Danelljan M, H?ger G, Khan F S, et al. Learning spatially regularized correlation filters for visual tracking . [C]∥2015 IEEE International Conference on Computer Vision (ICCV), December 7-13, 2015, Santiago, Chile. New York: IEEE. 2015, 4310-4318.

【3】Dong Q J, He X D, Ge H Y, et al. Adaptive merging complementary learners for visual tracking based on probabilistic model [J]. Laser & Optoelectronics Progress. 2019, 56(16): 161505.
董秋杰, 何雪东, 葛海燕, 等. 基于概率模型的自适应融合互补学习跟踪算法 [J]. 激光与光电子学进展. 2019, 56(16): 161505.

【4】He X D, Zhou S Z. Fast scale adaptive kernel correlation filtering algorithm for target tracking [J]. Laser & Optoelectronics Progress. 2018, 55(12): 121501.
何雪东, 周盛宗. 快速尺度自适应核相关滤波目标跟踪算法 [J]. 激光与光电子学进展. 2018, 55(12): 121501.

【5】Bolme D, 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, June 13-18, 2010, San Francisco, CA, USA. New York: IEEE. 2010, 2544-2550.

【6】Zhang Y Y. Object tracking based on similar background and color histogram in HSV color space [J]. Electronics Optics & Control. 2019, 26(4): 100-105.
张宇阳. 基于相似背景与HSV空间颜色直方图的目标跟踪 [J]. 电光与控制. 2019, 26(4): 100-105.

【7】Ma C, Huang J B, Yang X K, et al. Hierarchical convolutional features for visual tracking . [C]∥2015 IEEE International Conference on Computer Vision (ICCV), December 7-13, 2015, Santiago, Chile. New York: IEEE. 2015, 3074-3082.

【8】Danelljan M, Robinson A, Khan F S, et al. Beyond correlation filters: learning continuous convolution operators for visual tracking [M]. ∥Leibe B, Matas J, Sebe N, et al. Computer vision-ECCV 2016. Lecture notes in computer science. Cham: Springer. 2016, 9909: 472-488.

【9】Dalal N, Triggs B. Histograms of oriented gradients for human detection . [C]∥2005 IEEE Computer Society Conference on Computer Vision and Pattern Recognition (CVPR''''05), June 20-25, 2005, San Diego, CA, USA. New York: IEEE. 2005, 8588935.

【10】Danelljan M, Hager G, Khan F S, et al. Convolutional features for correlation filter based visual tracking . [C]∥2015 IEEE International Conference on Computer Vision Workshop (ICCVW), December 7-13, 2015, Santiago, Chile. New York: IEEE. 2015, 621-629.

【11】Bertinetto L, Valmadre J, Henriques J F, et al. Fully-convolutional Siamese networks for object tracking [M]. ∥Hua G, Jégou H. Computer vision-ECCV 2016 Workshops. Lecture notes in computer science. Cham: Springer. 2016, 9914: 850-865.

【12】Possegger H, Mauthner T, Bischof H. In defense of color-based model-free tracking . [C]∥2015 IEEE Conference on Computer Vision and Pattern Recognition (CVPR), June 7-12, 2015, Boston, MA, USA. New York: IEEE. 2015, 2113-2120.

【13】Mao N, Yang D D, Li Y, et al. Spatial regularization correlation filtering tracking via deformable diversity similarity [J]. Acta Optica Sinica. 2019, 39(4): 0415002.
毛宁, 杨德东, 李勇, 等. 基于形变多样相似性的空间正则化相关滤波跟踪 [J]. 光学学报. 2019, 39(4): 0415002.

【14】Yang D D, Mao N, Yang F C, et al. Improved SRDCF object tracking via the best-buddies similarity [J]. Optics and Precision Engineering. 2018, 26(2): 492-502.
杨德东, 毛宁, 杨福才, 等. 利用最佳伙伴相似性的改进空间正则化判别相关滤波目标跟踪 [J]. 光学精密工程. 2018, 26(2): 492-502.

【15】Li F, Tian C, Zuo W M, et al. Learning spatial-temporal regularized correlation filters for visual tracking . [C]∥2018 IEEE/CVF Conference on Computer Vision and Pattern Recognition, June 18-23, 2018, Salt Lake City, UT, USA. New York: IEEE. 2018, 4904-4913.

【16】Galoogahi H K, Sim T, Lucey S. Multi-channel correlation filters . [C]∥2013 IEEE International Conference on Computer Vision, December 1-8, 2013, Sydney, NSW, Australia. New York: IEEE. 2013, 3072-3079.

【17】Zhang H, Du Y, Ning S R, et al. Pedestrian detection method based on Faster RCNN [J]. Sensors and Microsystems. 2019, 38(2): 147-149, 153.
张汇, 杜煜, 宁淑荣, 等. 基于Faster RCNN的行人检测方法 [J]. 传感器与微系统. 2019, 38(2): 147-149, 153.

【18】Boyd S, Parikh N, Chu E, et al. Distributed optimization and statistical learning via the alternating direction method of multipliers [J]. Foundations and Trends ?in Machine Learning. 2010, 3(1): 1-122.

【19】Sherman J, Morrison W J. Adjustment of an inverse matrix corresponding to a change in one element of a given matrix [J]. The Annals of Mathematical Statistics. 1950, 21(1): 124-127.

【20】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.

【21】Bertinetto L, Valmadre J, Golodetz S, et al. Staple: complementary learners for real-time tracking . [C]∥2016 IEEE Conference on Computer Vision and Pattern Recognition (CVPR), June 27-30, 2016, Las Vegas, NV, USA. New York: IEEE. 2016, 1401-1409.

【22】Danelljan M, H?ger G, Khan F S, et al. Accurate scale estimation for robust visual tracking[M]. ∥Valstar M: , 2014.

【23】Galoogahi H K, Fagg A, Lucey S. Learning background-aware correlation filters for visual tracking . [C]∥2017 IEEE International Conference on Computer Vision (ICCV), October 22-29, 2017, Venice, Italy. New York: IEEE. 2017, 1144-1152.

引用该论文

Liu Mingming,Pei Dong,Liu Jü,Zhu Donghui,Sun Haoxiang. Filter Tracking Based on Time Regularization and Background-Aware[J]. Laser & Optoelectronics Progress, 2019, 56(23): 231503

刘明明,裴东,刘举,祝东辉,孙浩翔. 基于时间正则化及背景感知的滤波器跟踪[J]. 激光与光电子学进展, 2019, 56(23): 231503

您的浏览器不支持PDF插件,请使用最新的(Chrome/Fire Fox等)浏览器.或者您还可以点击此处下载该论文PDF