Please wait a minute...
浙江大学学报(工学版)
浙江大学学报(工学版)
计算机技术﹑电信技术     
结合排序向量SVM的视频跟踪
于慧敏, 曾雄
浙江大学 信息与电子工程学系,浙江 杭州 310027
Visual tracking combined with ranking vector SVM
YU Hui-min, ZENG Xiong
Department of Information Science and Electronic Engineering, Zhejiang University, Hangzhou 310027, China
 全文: PDF(1748 KB)   HTML
摘要：

针对真实视频场景中复杂的目标外观变化问题,提出新的结合排序向量SVM（RV-SVM）的单目标视频跟踪算法.基于压缩感知理论,利用稀疏测量矩阵压缩多尺度图像特征.采用Median-Flow跟踪算法作为预测器,并为RV-SVM构建训练数据集,使算法能够适应真实场景中遇到的目标遮挡、3D旋转和目标快速移动等复杂情况.通过在线学习RV-SVM算法,对候选位置集进行排序,找到目标的真实位置.对不同视频序列的测试结果表明：该方法可以在目标运动、旋转以及光照和尺度发生变化的情况下实现准确的跟踪.
Abstract:

A novel single object video tracking algorithm with ranking vector SVM (RV-SVM) was proposed for complex changes of object appearance in realistic scenarios. A sparse measurement matrix based on compressive sensing theory could compress the multi-scale image features. A Median-Flow tracker algorithm was used as a predictor and to construct training data sets for RV-SVM algorithm, so that the algorithm could adapt complex conditions like object occlusion, 3D rotation and fast object motion. The real position of target was determined through training the RV-SVM algorithm online and ranking the candidate position set. Results of tests on variant video sequences show that the algorithm can achieve stable tracking either the object is moving, rotating or the illumination and scale is changing.
出版日期: 2015-06-01
:  TN 911  
基金资助:

国家自然科学基金资助项目（61471321）；国家“973”重点基础研究发展规划资助项目(2012CB316400)；中兴通讯资助项目
作者简介: 于慧敏（1963—）,男,教授,博导,从事计算机视觉与模式识别的教学科研工作.E-mail: yhm2005@zju.edu.cn
服务　　
把本文推荐给朋友
加入引用管理器
E-mail Alert
RSS
作者相关文章　　

引用本文:

于慧敏, 曾雄. 结合排序向量SVM的视频跟踪[J]. 浙江大学学报(工学版), 10.3785/j.issn.1008-973X.2015.06.003.

YU Hui-min, ZENG Xiong. Visual tracking combined with ranking vector SVM. JOURNAL OF ZHEJIANG UNIVERSITY (ENGINEERING SCIENCE), 10.3785/j.issn.1008-973X.2015.06.003.

链接本文:

http://www.zjujournals.com/eng/CN/10.3785/j.issn.1008-973X.2015.06.003    或    http://www.zjujournals.com/eng/CN/Y2015/V49/I6/1015

［1］ GRABNER H, GRABNER M, BISCHOF H. Real-time tracking via on-line boosting ［C］ ∥ Proceedings of British Machine Vision Conference. Edinburgh: BMVC, 2006, 1(5): 6.
［2］ GRABNER H, LEISTNER C, BISCHOF H. Semi-supervised on-line boosting for robust tracking ［M］∥Proceedings of European Conference on Computer Vision. Berlin Heidelberg: Springer, 2008: 234-247.
［3］ ZHANG K, ZHANG L, YANG M H. Real-time compressive tracking ［M］∥Proceedings of European Conference on Computer Vision. Berlin Heidelberg: Springer, 2012: 864-877.
［4］ ZHANG K, SONG H. Real-time visual tracking via online weighted multiple instance learning ［J］. Pattern Recognition, 2013, 46(1): 397-411.
［5］ BABENKO B, YANG M H, BELONGIE S. Robust object tracking with online multiple instance learning ［J］. IEEE Transactions on Pattern Analysis and Machine Intelligence, 2011, 33(8): 1619-1632.
［6］ SANTNER J, LEISTNER C, SAFFARI A, et al. PROST: parallel robust online simple tracking ［C］ ∥ Proceedings of IEEE Conference on Computer Vision and Pattern Recognition. San Francisco: IEEE, 2010: 723-730.
［7］ BAI Y, TANG M. Robust visual tracking with ranking SVM ［C］∥Proceedings of IEEE Conference on Image Processing. Brussels: IEEE, 2011: 517-520.
［8］ COMANICIU D, RAMESH V, MEER P. Real-time tracking of non-rigid objects using mean shift ［C］∥Proceedings of IEEE Conference on Computer Vision and Pattern Recognition. Hilton Head: IEEE, 2000, 2: 142-149.
［9］ MEI X, LING H. Robust visual tracking and vehicle classification via sparse representation ［J］. IEEE Transactions on Pattern Analysis and Machine Intelligence, 2011, 33(11): 2259-2272.
［10］ ADAM A, RIVLIN E, SHIMSHONI I. Robust fragments-based tracking using the integral histogram ［C］ ∥ Proceedings of IEEE Conference on Computer Vision and Pattern Recognition. New York: IEEE, 2006: 798-805.
［11］ KWON J, LEE K M. Visual tracking decomposition ［C］ ∥ Proceedings of IEEE Conference on Computer Vision and Pattern Recognition. San Froncisco: IEEE, 2010: 1269-1276.
［12］ ROSS D, LIM J, LIN R S, et al. Incremental learning for robust visual tracking ［J］. International Journal of Computer Vision, 2008, 77(1-3): 125-141.
［13］ YU H, KIM J, KIM Y, et al. An efficient method for learning nonlinear ranking SVM functions ［J］. Information Sciences, 2012, 209(20): 37-48.
［14］ ZHANG K, ZHANG L, YANG M H. Fast compressive tracking ［J］. IEEE Transactions on Pattern Analysis and Machine Intelligence, 2014, 36(10): 2002-2015.
［15］ KALAL Z, MIKOLAJCZYK K, MATAS J. Forward-backward error: automatic detection of tracking failures ［C］ ∥ Proceedings of IEEE Conference on Pattern Recognition. San Froncisco: IEEE, 2010: 2756-2759.
［16］ KALAL Z, MIKOLAJCZYK K, MATAS J. Tracking-learning-detection ［J］. IEEE Transactions on Pattern Analysis and Machine Intelligence, 2012, 34(7): 1409-1422.
［17］ BOUGUET J Y. Pyramidal implementation of the Lucas Kanade feature tracker description of the algorithm, openCV documetation ［R］. Santa Clara, CA: Intel Corporation, Intel Microprocessor Research Labs. 1999.
［18］ HERBRICH R, GRAEPEL T, OBERMAYER K. Support vector learning for ordinal regression ［C］∥Proceedings of International Conference on Artificial Neural Networks. Edinburgh: ICANN, 1999: 97-102.
［19］ FREUND Y, IYER R, SCHAPIRE R E. An efficient boosting algorithm for combining preferences ［J］. Journal of Machine Learning Research, 2003, 4: 933-969.
［20］ BURGES C, SHAKED T, RENSHAW E, et al. Learning to rank using gradient descent ［C］∥Proceedings of International Conference on Machine Learning. Bonn: ICML, 2005: 89-96.
［21］ YANG P, LIU Q, METAXAS D N. RankBoost with L1 regularization for facial expression recognition and intensity estimation ［C］∥Proceedings of IEEE International Conference on Computer Vision. Kyoto: IEEE,2009: 1018-1025.
［22］ BAI Y, TANG M. Robust visual tracking via weakly supervised ranking SVM ［C］∥Proceedings of IEEE Conference on Computer Vision and Pattern Recognition. Providence: IEEE, 2012: 1854-1861.
［23］ GU S, ZHENG Y, TOMASI C. Efficient visual object tracking with online nearest neighbor classifier ［C］∥Proceedings of Asian Conference on Computer Vision. Queenstown: ACCV,2010: 271-282.
[1] 吴晨曦, 张旻, 王可人. 基于二级嵌套阵列的宽频段欠定波达方向估计[J]. 浙江大学学报(工学版), 2017, 51(5): 1016-1023.
[2] 谢罗峰, 徐慧宁, 黄沁元, 赵越, 殷国富. 应用双树复小波包和NCA-LSSVM检测磁瓦内部缺陷[J]. 浙江大学学报(工学版), 2017, 51(1): 184-191.
[3] 王志, 朱世强, 卜琰, 郭振民. 改进导向滤波器立体匹配算法[J]. 浙江大学学报(工学版), 2016, 50(12): 2262-2269.
[4] 陈阔, 冯华君, 徐之海, 李奇, 陈跃庭. 细节保持的快速曝光融合[J]. 浙江大学学报(工学版), 2015, 49(6): 1048-1054.
[5] 江燊煜,陈阔,徐之海,冯华君,李奇,陈跃庭. 基于曝光适度评价的多曝光图像融合方法[J]. 浙江大学学报(工学版), 2015, 49(3): 470-475.
[6] 朱株,刘济林. 基于马尔科夫随机场的三维激光雷达路面实时分割[J]. 浙江大学学报(工学版), 2015, 49(3): 464-469.
[7] 李江,赵雅琼,包晔华. 基于混沌和替代数据法的中风病人声音分析[J]. 浙江大学学报(工学版), 2015, 49(1): 36-41.
[8] 童基均, 张光磊, 蔡强, 简锦明,郭希山. 阈值随机共振及其在低质量浓度气体检测中的应用[J]. 浙江大学学报(工学版), 2015, 49(1): 15-19.
[9] 潘能杰,于慧敏. 边缘加强型的彩色空间最稳极值区域[J]. 浙江大学学报(工学版), 2014, 48(7): 1241-1247.
[10] 吴鹏洲,于慧敏,曾雄. 基于正则化风险最小化的目标计数[J]. 浙江大学学报(工学版), 2014, 48(7): 1226-1233.
[11] 岳克强, 孙玲玲, 游彬, 楼立恒. 基于欠定盲分离的并行识别防碰撞算法[J]. 浙江大学学报(工学版), 2014, 48(5): 865-870.
[12] 项楠, 赵航芳, 宫先仪. 非寻复域状态-空间滤波[J]. J4, 2014, 48(4): 727-733.
[13] 杨力, 朱株, 刘济林. 一种嵌入式汽车鸟瞰全景图拼接算法[J]. J4, 2014, 48(2): 292-296.
[14] 刘凤霞, 潘翔, 宫先仪. 螺旋线阵匹配场三维定位[J]. J4, 2013, 47(1): 62-69.
[15] 刘志坤, 刘忠, 付学志, 宁小玲. 改进的变步长自适应滤波及Eckart加权抑噪算法[J]. J4, 2012, 46(6): 1014-1020.