基于自适应组合核的鲁棒视频目标跟踪算法

doi:10.11772/j.issn.1001-9081.2018051139

计算机应用 ›› 2018, Vol. 38 ›› Issue (12): 3372-3379.DOI: 10.11772/j.issn.1001-9081.2018051139

基于自适应组合核的鲁棒视频目标跟踪算法

刘培强^1,2, 张加惠¹, 吴大伟³, 安志勇¹

1. 山东工商学院计算机科学与技术学院, 山东烟台 264005;
2. 山东省高等学校协同创新中心:未来智能计算, 山东烟台 264005;
3. 东北林业大学交通学院, 哈尔滨 150040

收稿日期:2018-06-01 修回日期:2018-08-14 出版日期:2018-12-10 发布日期:2018-12-15
通讯作者: 刘培强
作者简介:刘培强(1970-),男,山东淄博人,教授,博士,CCF会员,主要研究方向:计算机视觉、机器学习、目标跟踪、算法及其复杂性理论;张加惠(1995-),女,山东日照人,硕士研究生,主要研究方向:计算机视觉、目标跟踪;吴大伟(1999-),男,山西怀仁人,主要研究方向:智能交通;安志勇(1975-),男,山东烟台人,副教授,博士,主要研究方向:机器学习。
基金资助:
山东省自然科学基金资助项目（ZR2014FL007）；烟台市重点研发计划项目（2017ZH065）；赛尔网络下一代互联网技术创新项目（NGII20161204）。

Robust video object tracking algorithm based on self-adaptive compound kernel

LIU Peiqiang^1,2, ZHANG Jiahui¹, WU Dawei³, AN Zhiyong¹

1. School of Computer Science and Technology, Shandong Technology and Business University, Yantai Shandong 264005, China;
2. Co-Innovation Center of Shandong Colleges and Universities:Future Intelligent Computing, Yantai Shandong 264005, China;
3. School of Traffic, Northeast Forestry University, Harbin Heilongjiang 150040, China

Received:2018-06-01 Revised:2018-08-14 Online:2018-12-10 Published:2018-12-15
Contact: 刘培强
Supported by:
This work is partially supported by the Natural Science Foundation of Shandong Province (ZR2014FL007), the Yantai Key Research and Development Project (2017ZH065), the Next Generation Internet Technology Innovation Project of the CERNET (NGII20161204).

摘要/Abstract

摘要： 为了解决核化相关滤波器（KCF）在复杂场景下鲁棒性差的问题，提出了基于自适应组合核（SACK）的目标跟踪算法。跟踪任务分为位置跟踪和尺度跟踪两个独立部分。首先，以线性核和高斯核的自适应组合作为核跟踪滤波器，构造了SACK权重的风险目标函数。该函数根据核的响应值自适应调整线性核和高斯核权重，不仅考虑了不同核响应输出的经验风险泛函最小，而且考虑了极大响应值的风险泛函，同时具有局部核和全局核的优点。然后，根据该滤波器的输出响应得到目标精确位置，设计了基于目标极大响应值的自适应更新率，针对位置跟踪滤波器进行自适应更新。最后，利用尺度跟踪器对目标尺度进行估计。实验结果表明，所提算法的成功率和距离精度在OTB-50数据库表现最优，比KCF算法分别高6.8个百分点和4.1个百分点，比双向尺度估计跟踪（BSET）算法分别高2个百分点和3.2个百分点。该算法对形变和遮挡等复杂场景具有很强的适应能力。

关键词: 目标跟踪, 傅里叶变换, 核化相关滤波器, 组合核, 岭回归

Abstract: In order to solve the problem of poor robustness of Kernelized Correlation Filter (KCF) in complex scenes, a new object tracking algorithm based on Self-Adaptive Compound Kernel (SACK) was proposed. The tracking task was decomposed into two independent subtasks:position tracking and scale tracking. Firstly, the risk objective function of SACK weight was constructed by using the self-adaptive compound of linear kernel and Gaussian kernel as the kernel tracking filter. The weights of linear kernel and Gaussian kernel were adjusted adaptively by the constructed function according to the response values of kernels, which not only considered the minimum empirical risk function of different kernel response outputs, but also considered the risk function of maximum response value, and had the advantages of local kernel and global kernel. Then, the exact position of object was obtained according to the output response of the SACK filter, and the adaptive update rate based on the maximum response value of object was designed to adaptively update the position tracking filter. Finally, the scale tracker was used to estimate the object scale. The experimental results show that, the success rate and distance precision of the proposed algorithm are optimal on OTB-50 database, which is 6.8 percentage points and 4.1 percentage points higher than those of KCF algorithm respectively, 2 percentage points and 3.2 percentage points higher than those of Bidirectional Scale Estimation Tracker (BSET) algorithm respectively. The proposed algorithm has strong adaptability to complex scenes such as deformation and occlusion.

Key words: object tracking, Fourier transform, Kernel Correlation Filter (KCF), compound kernel, ridge regression

中图分类号:

TP391.4

刘培强, 张加惠, 吴大伟, 安志勇. 基于自适应组合核的鲁棒视频目标跟踪算法[J]. 计算机应用, 2018, 38(12): 3372-3379.

LIU Peiqiang, ZHANG Jiahui, WU Dawei, AN Zhiyong. Robust video object tracking algorithm based on self-adaptive compound kernel[J]. Journal of Computer Applications, 2018, 38(12): 3372-3379.

参考文献

[1] ZHAO Z Q, WANG T J, LIU F, et al. Remarkable local resampling based on particle filter for visual tracking[J]. Multimedia Tools an Applications, 2017, 76(1):835-860.
[2] HSIA C H, LIOU Y J, CHIANG J S. Directional prediction CamShift algorithm based on adaptive search pattern for moving object tracking[J]. Journal of Real-Time Image Processing, 2016, 12(1):183-195.
[3] VOJIR T, NOSKOVA J, MATAS J. Robust scale-adaptive mean-shift for tracking[C]//Proceedings of the 2013 Scandinavian Conference on Image Analysis, LNCS 7944. Berlin:Springer, 2013:652-663.
[4] JEONG J M, YOON T S, PARK J B. Mean shift tracker combined with online learning-based detector and kalman filtering for real-time tracking[J]. Expert Systems with Applications, 2017, 79:194-206.
[5] ZHANG L, VARADARAJAN J, SUGANTHAN P N, et al. Robust visual tracking using oblique random forests[C]//Proceedings of the 2017 IEEE Conference on Computer Vision and Pattern Recognition. Piscataway, NJ:IEEE, 2017:5825-5834.
[6] 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. Piscataway, NJ:IEEE, 2010:2544-2550.
[7] DANELLJAN M, HÄGER G, KHAN F S, et al. Discriminative scale space tracking[J]. Transactions on Pattern Analysis and Machine Intelligence, 2017, 39(8):1561-1575.
[8] ZHANG K H, ZHANG L, LIU Q S, et al. Fast visual tracking via spatio-temporal context learning[C]//Proceedings of the 2014 European Conference on Computer Vision. Berlin:Springer, 2014:127-141.
[9] MUELLER M, SMITH N, GHANEM B. Context-aware correlation filter tracking[C]//Proceedings of the 2017 IEEE Conference on Computer Vision and Pattern Recognition. Piscataway, NJ:IEEE, 2017:1387-1395.
[10] LIU T, WANG G, YANG Q X, et al. Part-based tracking via discriminative correlation filters[J/OL]. IEEE Transactions on Circuits & Systems for Video Technology, 2016.[2018-03-26]. https://ieeexplore.ieee.org/document/7779076/keywords#keywords.
[11] DANELLJAN M, HÄGER G, KHAN F S, et al. Learning spatially regularized correlation filters for visual tracking[C]//Proceedings of the 2015 IEEE International Conference on Computer Vision. Piscataway, NJ:IEEE,2015:4310-4318.
[12] 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.
[13] 朱明敏,胡茂海.基于相关滤波器的长时视觉目标跟踪方法[J].计算机应用,2017,37(5):1466-1470.(ZHU M M, HU M H. Long-term visual object tracking algorithm based on correlation filter[J]. Journal of Computer Applications, 2017, 37(5):1466-1470.)
[14] 钱堂慧,罗志清,李果家,等.核相关滤波跟踪算法的尺度自适应改进[J].计算机应用,2017,37(3):811-816.(QIAN T H, LUO Z Q, LI G J, et al. Scale adaptive improved of kernel correlation filter tracking algorithm[J]. Journal of Computer Applications, 2017, 37(3):811-816.)
[15] 李麒骥,李磊民,黄玉清.尺度自适应的核相关滤波跟踪器[J].计算机应用,2016,36(12):3385-3388.(LI Q J, LI L M, HUANG Y Q. Scale adaptive tracker based on kernelized correlation filtering[J]. Journal of Computer Applications, 2016, 36(12):3385-3388.)
[16] DANELLJAN M, HÄGER G, KHAN F S, et al. Adaptive decontamination of the training set:a unified formulation for discriminative visual tracking[C]//Proceedings of the 2016 IEEE Conference on Computer Vision and Pattern Recognition. Piscataway, NJ:IEEE, 2016:1430-1438.
[17] MA C, YANG X K, ZHANG C Y,et al. Long-term correlation tracking[C]//Proceedings of the 2015 IEEE Conference on Computer Vision and Pattern Recognition. Piscataway, NJ:IEEE, 2015:5388-5396.
[18] WANG M M, LIU Y, HUANG Z Y. Large margin object tracking with circulant feature maps[C]//Proceedings of the 2017 IEEE Conference on Computer Vision and Pattern Recognition. Piscataway, NJ:IEEE, 2017:4800-4808.
[19] NAM H S, HAN B Y. Learning multi-domain convolutional neural networks for visual tracking[C]//Proceedings of the 2016 IEEE Conference on Computer Vision and Pattern Recognition. Piscataway, NJ:IEEE, 2016:4293-4302.
[20] ZHANG K H, LIU Q S, WU Y, et al. Robust visual tracking via convolutional networks without training[J]. IEEE Transactions on Image Processing, 2016, 25(4):1779-1792.
[21] VALMADRE J, BERTINETTO L, HENRIQUES J. End-to-end representation learning for correlation filter based tracking[C]//Proceedings of the 2017 IEEE Conference on Computer Vision and Pattern Recognition. Piscataway, NJ:IEEE, 2017:5000-5008.
[22] 田中大,高宪文,石彤.用于混沌时间序列预测的组合核函数最小二乘支持向量机[J].物理学报,2014,63(16):66-76.(TIAN Z D, GAO X W, SHI T. Combination kernel function least squares support vector machine for chaotic time series predict[J]. Acta Physica Sinica, 2014, 63(16):66-76.)
[23] 张亚男,杨慧中.基于RVM组合核优化的软测量模型研究[J].系统仿真学报,2018,30(1):272-277.(ZHANG Y N, YANG H Z. RVM Soft sensing model based on optimized combined kernel function[J]. Journal of System Simulation, 2018, 30(1):272-277.)
[24] 夏红霞,丁子春,李哲,等.一种新的自适应组合核函数[J].武汉理工大学学报,2009,31(3):49-53.(XIA H X, DING Z C, LI Z, et al. A adaptive compound kernel function of support vector machines[J]. Journal of Wuhan University of Technology, 2009, 31(3):49-53.)
[25] WU Y, LIM J W, YANG M H,et al. Online object tracking:a benchmark[C]//Proceedings of the 2013 IEEE Conference on Computer Vision and Pattern Recognition. Piscataway, NJ:IEEE, 2013:2411-2418.
[26] AN Z Y, GUAN H, LI J J. Robust visual tracking using the bidirectional scale estimation[J]. Mathematical Problems in Engineering, 2017, 2017:Article ID 3276103.
[27] HARE S, SAFFARI A, TORR P H S. Struck:structured output tracking with kernels[C]//Proceedings of the 2011 International Conference on Computer Vision. Piscataway, NJ:IEEE, 2011:263-270.
[28] KALAL Z, MIKOLAJCZYK K, MATAS J. Tracking-learning-detection[J]. IEEE Transactions on Pattern Analysis and Machine Intelligence, 2012, 34(7):1409-1422.
[29] ORON S, BAR-HILLEL A, LEVI D, et al. Locally orderless tracking[C]//Proceedings of the 2012 IEEE Conference on Computer Vision and Pattern Recognition. Piscataway, NJ:IEEE, 2012:1940-1947.
[30] ZHANG K H, ZHANG L, YANG M H. Fast compressive tracking[J]. IEEE Transactions on Pattern Analysis and Machine Intelligence, 2014, 36(10):2002-2015.
[31] HE S F, LAU R W H, YANG Q X, et al. Robust object tracking via locality sensitive histograms[J]. IEEE Transactions on Circuits and Systems for Video Technology, 2017, 27(5):1006-1017.
[32] DANELLJAN M, KHAN F S, FELSBERG M, et al. Adaptive color attributes for real-time visual tracking[C]//Proceedings of the 2014 IEEE Conference on Computer Vision and Pattern Recognition. Piscataway, NJ:IEEE, 2014:1090-1097.

基于自适应组合核的鲁棒视频目标跟踪算法

Robust video object tracking algorithm based on self-adaptive compound kernel

PDF

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

[1]	姬张建, 任兴旺. 带旋转与尺度估计的全卷积孪生网络目标跟踪算法[J]. 计算机应用, 2021, 41(9): 2705-2711.
[2]	徐丽云, 闫涛, 钱宇华. 基于级联混沌系统的分数域语音加密算法[J]. 计算机应用, 2021, 41(9): 2623-2630.
[3]	钟莎, 黄玉清. 基于孪生区域候选网络的无人机指定目标跟踪[J]. 计算机应用, 2021, 41(2): 523-529.
[4]	汪志远, 降爱莲, 奥斯曼·穆罕默德. 基于正则互表示的无监督特征选择方法[J]. 计算机应用, 2020, 40(7): 1896-1900.
[5]	张杰, 常天庆, 戴文君, 郭理彬, 张雷. 基于相关滤波与颜色概率模型的目标跟踪算法[J]. 计算机应用, 2020, 40(6): 1774-1782.
[6]	王杰, 王让定, 严迪群, 林昱臻. 基于卷积神经网络框架的回声隐藏检测方法[J]. 计算机应用, 2020, 40(2): 375-380.
[7]	何祥宇, 李静, 杨数强, 夏玉杰. 基于ET-PHD滤波器和变分贝叶斯近似的扩展目标跟踪算法[J]. 计算机应用, 2020, 40(12): 3701-3706.
[8]	石国强, 赵霞. 基于联合优化的强耦合孪生区域推荐网络的目标跟踪算法[J]. 计算机应用, 2020, 40(10): 2822-2830.
[9]	杨博, 蔺素珍, 禄晓飞, 李大威, 秦品乐, 左健宏. 基于多特征融合与分层数据关联的空中红外多目标跟踪方法[J]. 计算机应用, 2020, 40(10): 3075-3080.
[10]	何祥宇, 于斌, 夏玉杰. 粒子概率假设密度平滑器异常平滑问题的解决方法[J]. 计算机应用, 2020, 40(1): 299-303.
[11]	俞皓芳, 孙力帆, 付主木. 基于改进K-means++聚类的多扩展目标跟踪算法[J]. 计算机应用, 2020, 40(1): 271-277.
[12]	万永菁, 王博玮, 娄定风. 基于三维卷积神经网络的虫音特征识别方法[J]. 计算机应用, 2019, 39(9): 2744-2748.
[13]	贾永超, 何小卫, 郑忠龙. 融合重检测机制的卷积回归网络目标跟踪算法[J]. 计算机应用, 2019, 39(8): 2247-2251.
[14]	韩玉兰, 韩崇昭. 非线性系统的多扩展目标跟踪算法[J]. 计算机应用, 2019, 39(5): 1318-1324.
[15]	杨春德, 刘京, 瞿中. 分离窗口快速尺度自适应目标跟踪算法[J]. 计算机应用, 2019, 39(4): 1145-1149.