基于双向稀疏表示的鲁棒目标跟踪算法

王保宪; 赵保军; 唐林波; 王水根; 吴京辉

doi:10.7498/aps.63.234201

摘要

目前, 基于稀疏表示的目标跟踪通常为在目标模板集上重构候选样本的正向模型或者在候选样本集上描述目标模板的反向模型. 两个模型的共同点是均需计算候选样本与模板集合之间的稀疏相关系数矩阵. 基于此, 建立了一个双向联合稀疏表示的跟踪模型, 该模型通过L2范数约束正反向稀疏相关系数矩阵达到一致收敛. 与之前的单向稀疏表示模型相比, 双向稀疏表示跟踪模型在正反向联合求解框架下可以更加充分地挖掘所有候选样本与模板集之间的稀疏映射关系, 并将稀疏映射表上对正负模板区分度最好的候选样本作为目标. 基于加速逼近梯度(accelerated proximal gradient)快速算法, 以矩阵形式推导了双向稀疏表示模型的求解框架, 使得候选样本集和目标模板集均以矩阵方式并行求解, 在一定程度上提高了计算效率. 实验数据表明所提出的算法优于传统的单向稀疏表示目标跟踪算法.

关键词:

Abstract

At present the visual tracking model based on sparse representation is mainly divided into two types: one is to use the template set to reconstruct candidate samples, which is called forward model; the other is to project the template set into a candidate space, which is called reverse model. What the two models have in common is to compute the sparse correlation coefficient matrix of candidate sample and template set. Based on this, the paper establishes a bidirectional cooperative sparse representation tracking model. Using L2-norm constraint item, the forward and reverse sparse correlation matrix coefficients could be uniformly convergent. In comparison to conventional unidirectional sparse tracking model, bidirectional sparse tracking model could fully excavate the sparse mapping relation of the whole candidate sample and template set. And the candidate that scores highest in the sparse mapping table for the positive and negative templates is the tracking result. Based on the accelerated proximal gradient fast method, the paper derives the optimum solution (in matrix form) of bidirectional sparse tracking model. As a result, it allows the candidates and templates to be calculated in parallel, which can improve the calculation efficiency to some extent. Numerical examples show that the proposed tracking algorithm has certain priority over against the conventional unidirectional sparse tracking methods.

Keywords:

作者及机构信息

1.
北京理工大学信息与电子学院, 北京 100081

基金项目: 国家高技术研究发展计划(批准号:2012AA8012011C)资助的课题.

Authors and contacts

1.
School of Information and Electronics, Beijing Institute of Technology, Beijing 100081, China

Funds: Project supported by the National High Technology Research and Development Program of China (Grant No. 2012AA8012011C).

参考文献

[1]	Gao W, Tang Y, Zhu M 2014 Acta Phys. Sin. 63 094204 (in Chinese) [高文, 汤洋, 朱明 2014 物理学报 63 094204]
[2]	Donoho D L 2006 IEEE Trans. Inform. Theory 52 1289
[3]	Wright J, Yang A Y, Ganesh A, Sastry S S, Ma Y 2009 IEEE Trans. Pattern Anal. Mach. Intell. 31 210
[4]	Deng C Z, Tian W, Chen P, Wang S Q, Zhu H S, Hu S F 2014 Acta Phys. Sin. 63 044202 (in Chinese) [邓承志, 田伟, 陈盼, 汪胜前, 朱华生, 胡赛凤 2014 物理学报 63 044202]
[5]	Song C X, Ma K, Qin C, Xiao P 2013 Acta Phys. Sin. 62 040702 (in Chinese) [宋长新, 马克, 秦川, 肖鹏 2013 物理学报 62 040702]
[6]	Mei X, Ling H B 2011 IEEE Trans. Pattern Anal. Mach. Intell. 33 2259
[7]	Bao C L, Wu Y, Ling H B, Ji H 2012 Proceedings of IEEE Conference on Computer Vision and Pattern Recognition Providence, USA, June 16-21, 2012 p1830
[8]	Zhong W, Lu H C, Yang M H 2012 Proceedings of IEEE Conference on Computer Vision and Pattern Recognition Providence, USA, June 16-21, 2012 p1838
[9]	Liu H P, Sun F C 2010 Proceedings of IEEE 20th Conference on Pattern Recognition Istanbul, Turkey, August 23-26, 2010 p1702
[10]	Zhuang B H, Lu H C, Xiao Z Y, Wang D 2014 IEEE Trans. Image Proc. 23 1872
[11]	Li X, Hu W M, Shen C H, Zhang Z F, Dick A 2013 ACM Trans. Intell. Syst. Technol. 4 58
[12]	Toh K C, Yun S 2010 Pac. J. Optim. 6 615
[13]	Zhang K H, Zhang L, Yang M H 2012 Proceedings of European Conference on Computer Vision Berlin, Germany, October 7-13, 2012 p864
[14]	Babenko B, Yang M H, Belongie S 2011 IEEE Trans. Pattern Anal. Mach. Intell. 33 1619
[15]	Wu Y, Lim J, Yang M H 2013 Proceedings of IEEE Conference on Computer Vision and Pattern Recognition Portland, USA, June 23-28, 2013 p2411

施引文献

[1]	Gao W, Tang Y, Zhu M 2014 Acta Phys. Sin. 63 094204 (in Chinese) [高文, 汤洋, 朱明 2014 物理学报 63 094204]
[2]	Donoho D L 2006 IEEE Trans. Inform. Theory 52 1289
[3]	Wright J, Yang A Y, Ganesh A, Sastry S S, Ma Y 2009 IEEE Trans. Pattern Anal. Mach. Intell. 31 210
[4]	Deng C Z, Tian W, Chen P, Wang S Q, Zhu H S, Hu S F 2014 Acta Phys. Sin. 63 044202 (in Chinese) [邓承志, 田伟, 陈盼, 汪胜前, 朱华生, 胡赛凤 2014 物理学报 63 044202]
[5]	Song C X, Ma K, Qin C, Xiao P 2013 Acta Phys. Sin. 62 040702 (in Chinese) [宋长新, 马克, 秦川, 肖鹏 2013 物理学报 62 040702]
[6]	Mei X, Ling H B 2011 IEEE Trans. Pattern Anal. Mach. Intell. 33 2259
[7]	Bao C L, Wu Y, Ling H B, Ji H 2012 Proceedings of IEEE Conference on Computer Vision and Pattern Recognition Providence, USA, June 16-21, 2012 p1830
[8]	Zhong W, Lu H C, Yang M H 2012 Proceedings of IEEE Conference on Computer Vision and Pattern Recognition Providence, USA, June 16-21, 2012 p1838
[9]	Liu H P, Sun F C 2010 Proceedings of IEEE 20th Conference on Pattern Recognition Istanbul, Turkey, August 23-26, 2010 p1702
[10]	Zhuang B H, Lu H C, Xiao Z Y, Wang D 2014 IEEE Trans. Image Proc. 23 1872
[11]	Li X, Hu W M, Shen C H, Zhang Z F, Dick A 2013 ACM Trans. Intell. Syst. Technol. 4 58
[12]	Toh K C, Yun S 2010 Pac. J. Optim. 6 615
[13]	Zhang K H, Zhang L, Yang M H 2012 Proceedings of European Conference on Computer Vision Berlin, Germany, October 7-13, 2012 p864
[14]	Babenko B, Yang M H, Belongie S 2011 IEEE Trans. Pattern Anal. Mach. Intell. 33 1619
[15]	Wu Y, Lim J, Yang M H 2013 Proceedings of IEEE Conference on Computer Vision and Pattern Recognition Portland, USA, June 23-28, 2013 p2411

[1]	伍静, 崔春凤, 欧阳滔, 唐超. 基于贝叶斯算法的5-7环缺陷石墨烯纳米带热电性能优化设计. 物理学报, 2023, 72(4): 047201. doi: 10.7498/aps.72.20222135
[2]	郝望, 段睿, 杨坤德. 联合简正波水波和底波频散特性的贝叶斯地声参数反演. 物理学报, 2023, 72(5): 054303. doi: 10.7498/aps.72.20221717
[3]	娄月申, 郭文军. 贝叶斯深度神经网络对于核质量预测的研究. 物理学报, 2022, 71(10): 102101. doi: 10.7498/aps.71.20212387
[4]	杨棣, 王元美, 李军刚. 贝叶斯频率估计中频率的先验分布对有色噪声作用的影响. 物理学报, 2018, 67(6): 060301. doi: 10.7498/aps.67.20171911
[5]	张羽, 罗秀娟, 曹蓓, 陈明徕, 刘辉, 夏爱利, 兰富洋. 傅里叶望远镜发射阵列的冗余度及冗余度-斯特列尔比-目标信息特性分析. 物理学报, 2016, 65(11): 114201. doi: 10.7498/aps.65.114201
[6]	李倩倩, 阳凡林, 张凯, 郑炳祥. 不确定海洋环境中基于贝叶斯理论的声源运动参数估计方法. 物理学报, 2016, 65(16): 164304. doi: 10.7498/aps.65.164304
[7]	郭苗苗, 王昱婧, 徐桂芝, Griffin Milsap, Nitish V. Thakor, Nathan Crone. 时变动态贝叶斯网络模型及其在皮层脑电网络连接中的应用. 物理学报, 2016, 65(3): 038702. doi: 10.7498/aps.65.038702
[8]	李慧, 赵琳, 李亮. 基于贝叶斯压缩感知的周跳探测与修复方法. 物理学报, 2016, 65(24): 249101. doi: 10.7498/aps.65.249101
[9]	张文杰, 王世元, 冯亚丽, 冯久超. 基于Huber的高阶容积卡尔曼跟踪算法. 物理学报, 2016, 65(8): 088401. doi: 10.7498/aps.65.088401
[10]	尹诗白, 王卫星, 王一斌, 李大鹏, 邓箴. 贝叶斯迭代联合双边滤波的散焦图像快速盲复原. 物理学报, 2016, 65(23): 234202. doi: 10.7498/aps.65.234202
[11]	陈典兵, 朱明, 高文, 王慧利, 杨航. 基于残差矩阵估计的稀疏表示目标跟踪算法. 物理学报, 2016, 65(19): 194201. doi: 10.7498/aps.65.194201
[12]	吴昊, 陈树新, 杨宾峰, 陈坤. 基于广义M估计的鲁棒容积卡尔曼滤波目标跟踪算法. 物理学报, 2015, 64(21): 218401. doi: 10.7498/aps.64.218401
[13]	马鸽, 胡跃明, 高红霞, 李致富, 郭琪伟. 基于物理总能量目标函数的稀疏重建模型. 物理学报, 2015, 64(20): 204202. doi: 10.7498/aps.64.204202
[14]	文方青, 张弓, 贲德. 基于块稀疏贝叶斯学习的多任务压缩感知重构算法. 物理学报, 2015, 64(7): 070201. doi: 10.7498/aps.64.070201
[15]	逯志宇, 王大鸣, 王建辉, 王跃. 基于时频差的正交容积卡尔曼滤波跟踪算法. 物理学报, 2015, 64(15): 150502. doi: 10.7498/aps.64.150502
[16]	高文, 汤洋, 朱明. 目标跟踪中目标模型更新问题的半监督学习算法研究. 物理学报, 2015, 64(1): 014205. doi: 10.7498/aps.64.014205
[17]	高文, 汤洋, 朱明. 复杂背景下目标检测的级联分类器算法研究. 物理学报, 2014, 63(9): 094204. doi: 10.7498/aps.63.094204
[18]	王娇, 周云辉, 黄玉清, 江虹. 基于贝叶斯网络的认知引擎设计与重配置. 物理学报, 2013, 62(3): 038402. doi: 10.7498/aps.62.038402
[19]	颜鹏程, 侯威, 钱忠华, 何文平, 孙建安. 基于贝叶斯理论的全球海温异常对500 hPa 温度场的影响分析. 物理学报, 2012, 61(13): 139202. doi: 10.7498/aps.61.139202
[20]	郝崇清, 王江, 邓斌, 魏熙乐. 基于稀疏贝叶斯学习的复杂网络拓扑估计. 物理学报, 2012, 61(14): 148901. doi: 10.7498/aps.61.148901

计量

文章访问数: 5580
PDF下载量: 721
被引次数: 0

姓名
邮箱
手机号码
标题
留言内容
验证码

搜索

留言板