基于特征显著性的均值漂移鲁棒目标跟踪

摘要/Abstract

摘要：

在均值漂移算法框架下，提出基于目标显著性的特征融合与在线模板更新策略，实现复杂动态环境下的鲁棒跟踪.通过目标区域与背景区域的特征对比定义了特征显著性测度.提出了基于特征显著性的色彩空间选择以及基于Gabor小波稀疏编码的纹理特征提取算法.通过特征显著性加权实现参考直方图模板的初始化，并在此基础上针对遮挡现象与目标自身形变的区别设计了在线模板更新策略.实验结果表明，本文方法与其他跟踪算法相比具有较强的鲁棒性和较高的准确性.

关键词: 目标跟踪, 均值漂移, 特征显著性, Gabor小波, 稀疏编码

Abstract:

This paper proposes the saliency-based feature fusion and template updating strategy in the meanshift framework to realize robust tracking in dynamic complex environment. Firstly, A feature saliency measurement is defined as the contrast between object and background. The optimal color space based on feature saliency and the texture feature based on Gabor wavelet sparse coding are extracted. The reference histogram template initialization is weighted by feature saliency and an online template updating strategy is presented for occlusion and object deformation. Experimental results show that the proposed model is more precise and more robust compared with up-to-date competing models.

Key words: object tracking, meanshift, feature saliency, Gabor wavelet, sparse coding

中图分类号:

TP 391

陈东岳，陈宗文. 基于特征显著性的均值漂移鲁棒目标跟踪[J]. 上海交通大学学报（自然版）, 2013, 47(11): 1807-1812.

CHEN Dongyue,CHEN Zongwen. Meanshift Robust Object Tracking Based on Feature Saliency[J]. Journal of Shanghai Jiaotong University, 2013, 47(11): 1807-1812.

参考文献

［1］Ta D N, Chen WeiChao, Gelfand N, et al. SURFTrac: Efficient tracking and continuous object recognition using local feature descriptors［C］∥IEEE Conf on Computer Vision and Pattern Recognition. Miami, LF: IEEE Press, 2009. 29372944.

［2］Jean J H, Lian FengLi. Robust visual servo control of a mobile robot for object tracking using shape parameters［J］. IEEE Transactions on Control Systems Technology, 2012, 20(6): 14611472.

［3］Wang Shu, Lu Huchuan, Fan Yang, et al. Superpixel tracking［C］∥13th IEEE International Conference on Computer Vision. Barcelona:［s.n.］, 2011: 13231330.

［4］Senst T, Evangelio R H, Sikora T. Detecting people carrying objects based on an optical flow motion model［C］∥2011 IEEE Workshop on Applications of Computer Vision. Kona, HI:［s.n.］, 2011: 301306.

［5］Grabner H, Bischof H. Online boosting and vision［C］∥Computer Society Conference on Computer Vision and Pattern Recognition. New York, USA:［s.n.］ 2006: 260267.

［6］Grabner H, Grabner M, Bischof H. Realtime tracking via online boosting［C］∥British Machine Vision Conference. Edinburgh, UK:［s.n.］, 2006: 4756.

［7］Babenko B, Yang MingHsuan, Belongie S. Robust object tracking with online multiple instance learning［J］. IEEE Transactions on Pattern Analysis and Machine Intelligence, 2011, 33(8): 16191632.

［8］Kwon J, Lee K. Visual tracking decomposition［C］∥IEEE Computer Society Conference on Computer Vision and Pattern Recognition. San Francisco, USA:［s.n.］, 2010: 12691276.

［9］Rui Yong, Chen YunQiang. Better proposal distributions: Object tracking using unscented particle filter［C］∥IEEE Computer Society Conference on Computer Vision and Pattern Recognition. Redmond, USA:［s.n.］，2001: 786793.

［10］Nummiaro K, KollerMeier E, Van Gool L. Object tracking with an adaptive colorbased particle filter［C］∥Proc of the 24th DAGM Symposium on Pattern Recognition. Zurich, Switzerland:［s.n.］, 2002: 353360.

［11］Tian Jun, Qian JianSheng, Li ShiYin, et al. Visual tracking with adaptive multicue fusion particle filter［J］. Optics and Precision Engineering, 2010, 18(10): 22542261(in Chinese).

［12］Dore A, Soto M, Regazzoni C S. Bayesian tracking for video analytics［J］. Signal Processing Magazine, 2010, 27(5): 4655.

［13］Comaniciu D, Ramesh V, Meer P. Realtime tracking of nonrigid objects using mean shift［C］∥IEEE Computer Society Conference on Computer Vision and Pattern Recognition. Hilton Head, USA:［s.n.］, 2000: 142149.

［14］Yang ChangJiang, Duraiswami R, Davis L. Efficient meanshift tracking via a new similarity measure［C］∥IEEE Computer Society Conference on Computer Vision and Pattern Recognition, College Park, USA:［s.n.］, 2005: 176183.

［15］Leichter I, Lindenbaum M, Rivlin E. MeanShift tracking with multiple reference color histograms［J］. Computer vision and Image Understanding, 2010, 114(3): 400408.

［16］Techmer A. Contourbased motion estimation and object tracking for realtime applications［C］∥2001 International Conference on Image Processing. Thessaloniki, Laurin:［s.n.］, 2001: 648651.

［17］Chen Qiang, Sun QuanSen, Heng Pheng Ann, et al. Twostage object tracking method based on kernel and active contour［J］. IEEE Transactions on Circuit and Systems for Video Technology, 2010, 20(4): 605609.

［18］Yu Lei, He ZhongShi, Cao Qi. Gabor texture representation method for face recognition using the Gamma and generalized Gaussian models［J］. Image and Vision Computing, 2010, 28(1): 177187.

［19］Mei Xue, Ling HaiBin. Robust visual tracking using ?1 minimization［C］∥12th IEEE International Conference on Computer Vision. Kyoto:［s.n.］, 2009: 14361443.

[1]	何清林, 赵宏宇, 陈雨轩, 侯翔昊. 基于CRLB的空中目标水下分布式声跟踪算法设计[J]. 空天防御, 2025, 8(1): 86-94.
[2]	李克勇, 李楚晨, 蔡云泽. 群对群拦截协同态势感知关键技术[J]. 空天防御, 2025, 8(1): 1-9.
[3]	唐胜景, 王太岩, 赵刚练, 郭杰, 李佳丽, 尹航. 面向目标跟踪的多传感器数据融合研究综述[J]. 空天防御, 2024, 7(4): 18-29.
[4]	郝亮, 黄颖浩, 姚莉秀, 蔡云泽. 一类基于模糊推理的具有机动自适应的目标跟踪算法[J]. 上海交通大学学报, 2024, 58(4): 468-480.
[5]	陈坤1, 2, 赵旭1, 董春玉1, 邸子超1, 陈宗枝1. 基于滤波器预测的抗遮挡目标跟踪算法[J]. J Shanghai Jiaotong Univ Sci, 2024, 29(3): 400-413.
[6]	刘增敏1, 2, 3, 4, 6, 王申涛5, 姚莉秀1, 2, 3, 蔡云泽1, 2, 3, 4, 6. 基于目标检测和特征提取网络的运动无人机平台下多目标跟踪[J]. J Shanghai Jiaotong Univ Sci, 2024, 29(3): 388-399.
[7]	. 基于多视角道路相机的时空关联三维车辆检测与跟踪系统[J]. J Shanghai Jiaotong Univ Sci, 2023, 28(1): 52-60.
[8]	何子述, 苏洋, 程婷. 基于共址MIMO雷达组网系统的机会协同目标跟踪方法[J]. 空天防御, 2022, 5(1): 6-11.
[9]	侯煜冠, 韩远鹏, 谢金月, 毛兴鹏. 基于SDE模型的雷达组目标角度跟踪方法[J]. 空天防御, 2021, 4(1): 41-46.
[10]	陈晓,李亚安,李余兴,蔚婧. 基于距离加权的概率数据关联机动目标跟踪算法[J]. 上海交通大学学报, 2018, 52(4): 474-479.
[11]	李坦坦，雷明. 基于启发式逻辑的概率假设密度滤波高效航迹起始方法[J]. 上海交通大学学报, 2018, 52(1): 63-69.
[12]	梁荔, 敬忠良, 董鹏, 李旻哲. 基于噪声方差估计的高斯混合概率假设密度滤波算法[J]. 上海交通大学学报, 2016, 50(9): 1355-1361.
[13]	李家强1a,2，赵春艳1b,2，陈金立1a,1b，赵荣华1b,2. WGS-84坐标系下改进的广义最小二乘配准算法[J]. 上海交通大学学报（自然版）, 2016, 50(05): 771-775.
[14]	吴迪, 唐勇奇, 万琴. 基于视觉场景复杂度多特征自适应融合的目标跟踪[J]. 上海交通大学学报, 2015, 49(12): 1868-1875.
[15]	王震, 敬忠良, 雷明, 秦彦源, 董鹏. 基于PHD的多扩展目标联合检测、跟踪与分类算法[J]. 上海交通大学学报, 2015, 49(11): 1589-1596.