Visual object tracking combined normal hedge and kernel sparse representation classification

KUANG Jin-jun; CHAI Yi; XIONG Qing-yu

doi:10.3788/OPE.20122011.2540

您当前的位置：

首页 >

文章列表页 >

Visual object tracking combined normal hedge and kernel sparse representation classification

Article | 更新时间：2020-08-12

- Visual object tracking combined normal hedge and kernel sparse representation classification
- Optics and Precision Engineering Vol. 20, Issue 11, Pages: 2540-2547(2012)
- 作者机构：
  
  重庆大学自动化学院重庆,400030
- 作者简介：
- 基金信息：
- DOI：10.3788/OPE.20122011.2540
  CLC： TP391
- Received：04 May 2012，
  
  Revised：02 July 2012，
  
  Published：10 November 2012
- 稿件说明：
移动端阅览
匡金骏, 柴毅, 熊庆宇. 结合标准对冲与核函数稀疏分类的目标跟踪[J]. 光学精密工程, 2012,20(11): 2540-2547

KUANG Jin-jun, CHAI Yi, XIONG Qing-yu. Visual object tracking combined normal hedge and kernel sparse representation classification[J]. Editorial Office of Optics and Precision Engineering, 2012,20(11): 2540-2547
匡金骏, 柴毅, 熊庆宇. 结合标准对冲与核函数稀疏分类的目标跟踪[J]. 光学精密工程, 2012,20(11): 2540-2547 DOI： 10.3788/OPE.20122011.2540.

KUANG Jin-jun, CHAI Yi, XIONG Qing-yu. Visual object tracking combined normal hedge and kernel sparse representation classification[J]. Editorial Office of Optics and Precision Engineering, 2012,20(11): 2540-2547 DOI： 10.3788/OPE.20122011.2540.

摘要

针对经典稀疏分类目标跟踪算法在噪声

遮挡等恶劣环境下精度不高的问题

提出了一种新的目标跟踪算法。该算法在标准对冲框架下结合了核函数稀疏分类方法以及自适应字典更新方法

能够较好地适应类间相似度较高与目标外形变化较大等恶劣情况。核函数技巧能够增强分类器性能

但通用方法求解凸优化问题的效率较低

不能满足目标跟踪问题的实时性要求

故提出用核函数随机坐标下降(KRCD)算法来高效求解稀疏系数

并使用核函数稀疏分类方法(KRCD-SRC)来计算各个粒子的代价值。为了避免模板漂移问题

解释了目标字典和背景字典的在线更新方法。最后

结合标准对冲算法估算目标的状态信息。在使用50个粒子进行跟踪时

本文算法的处理帧率能够达到14 frame/s。相比其它几种经典目标跟踪算法

本文算法具有更好的精确性和鲁棒性。

Abstract

To achieve the robust tracking for a visual object under challenging conditions in the noisy

occlusion and the deformation

a novel visual object tracking method is proposed in this paper. By combining the Kernel Sparse Representation Classification (KSRC) and adaptive dictionary updating method under Normal Hedge framework

this method can handle tough situations like high inter-class similarities and drastically target appearance variations. Although the KSRC enhances classification performance

standard convex optimization is not fast enough for tracking in real time. Thus an efficient Kernel Random Coordinate Descent(KRCD) method is proposed to calculate the sparse coefficient vector

and the KRCD-SRC classification method is taken to calculate the loss value of each particle. In order to avoid the template drifting

the adaptive dictionary updating method is also given. At last

the states of the target are estimated by the Normal Hedge. Experiments show that the average computing frame rate of the proposed method is 14 frame/s when 50 particles are used. Extensive test results suggest that the proposed method outperforms several state-of-art tracking methods in many complex conditions.

关键词

Keywords

references

JAWARD M, BULL D, CANAGARAJAH N. Sequential monte carlo methods for contour tracking of contaminant clouds[J]. Signal Processing, 2010, 90(1):249-260.[2] CHEN Q, SUN Q, HENG P A, XIA D S, et al.. Two-stage object tracking method based on kernel and active contour[J]. IEEE Transactions on Circuits and Systems for Video Technology, 2010, 20(4):605-609.[3] JING L, VADAKKEPAT P. Interacting mcmc particle filter for tracking maneuvering target[J]. Digital Signal Processing, 2010, 20(2):561-574.[4] HAN Z, JIAO J, ZHANG B, et al.. Visual object tracking via sample-based adaptive sparse representation (AdaSR)[J]. Pattern Recognition, 2011, 44(9):2170-2183.[5] XUE M, LING H B. Robust visual tracking and vehicle classification via sparse representation [J]. IEEE Transactions on Pattern Analysis and Machine Intelligence, 2011, 33(11):2259-2272.[6] WANG Q, CHEN F, XU X L, et al.. Online discriminative object tracking with local sparse representation. IEEE Workshop on Applications of Computer Vision (WACV), 2012:425-432.[7] RIGAMONTI R, BROWN M A, LEPETIT V. Are sparse representations really relevant for image classification?. IEEE Conference on Computer Vision and Pattern Recognition (CVPR), 2011:1545-1552.[8] ZHANG L, YANG M, FENG X CH. Sparse representation or collaborative representation:Which helps face recognition?[J]. IEEE International Conference on Computer Vision, 2011:471-478.[9] HUANG K, AVIYENTE S. Sparse representation for signal classification [J]. Advances in neural information processing, 2007, 16(15): 609-617.[10] GAO SH G, IVOR W T, CHIA L T. Kernel sparse representation for image classification and face recognition [J]. Lecture Notes in Computer Science, 2010, 63(14): 1-14.[11] KANG C C, LIAO SH C, XIANG SH M, et al.. Kernel sparse representation with local patterns for face recognition. 18th IEEE International Conference on Image Processing, 2011:3009-3012.[12] 龚俊亮,何昕,魏仲慧,等. 采用改进辅助粒子滤波的红外多目标跟踪 [J]. 光学精密工程, 2012, 20(2):413-421. GONG J L, HE X, WEI ZH H, et al.. Multiple infrared target tracking using improved auxiliary particle filter [J]. Opt. Precision Eng., 2012, 20(2):413-421. (in Chinese)[13] 吴川,杨冬,郝志成. 基于粒子滤波的彩色图像跟踪 [J]. 光学精密工程,2009, 17(10):2542-2547. WU CH, YANG D, HAO ZH CH.Color image tracking algorithm based on particle filter [J]. Opt. Precision Eng., 2009, 17(10):2542-2547. (in Chinese)[14] LI R, MAO L, ZHANG J R. Particle filter resampling based on optimized combinatorial algorithm. International Symposium on IT in Medicine and Education (ITME), 2011, 2:27-30.[15] CHAUDHURI K, FREUND Y, HSU D, A parameter-free hedging algorithm. Advances in Neural Information Processing Systems, 2009: 297-305.[16] ETTINGER E, FREUND Y. Particle Filtering on the Audio Localization Manifold, Arxiv preprint arXiv:1003.0659, 2010.[17] DONOHO D L.For most large underdetermined systems of linear equations the minimal l1-norm solution is also the sparsest solution. Communications on Pure and Applied Mathematics [J]. 2006, 59(6):797-829.[18] NESTEROV Y. Efficiency of coordinate descent methods on huge-scale optimization problems [J]. SIAM Journal on Optimization, 2012, 22(2):341-32.[19] PETER RICHTRIK, MARTIN TAK, Iteration complexity of randomized block-coordinate descent methods for minimizing a composite function. Arxiv preprint arXiv:1107.2848, 2011.[20] 朱明清,王智灵, 陈宗海. 基于灰色预测模型和粒子滤波的视觉目标跟踪算法[J]. 控制与决策, 2012, 27(1):53-57. ZHU M Q, WANG ZH L, CHEN Z H. Visual tracking algorithm based on grey prediction model and particle filter[J]. Control and Decision, 2012, 27(1):53-57. (in Chinese)

Views

404

下载量

CSCD

Alert me when the article has been cited

提交

Tools

Publicity Resources

A Transformer-based visual tracker via knowledge distillation

Special attribute-based cross-modal interactive fusion network for RGBT tracking

Salient region weighted correlation filter for object tracking

Correlation filter tracking algorithm based on double model

Related Author

JIA Xingxuan

LIU Mengqiao

LI Na

PAN Jinting

HUANG Kai

SHAO Xiaoqiang

LI Hao

LÜ Zhiyue

Related Institution

School of Communication and Information Engineering， Xi’an University of Posts and Telecommunications

College of Electrical and Control Engineering， Xi’an University of Science and Technology

University of Chinese Academy of Sciences

Changchun Institute of Optics， Fine Mechanics and Physics Chinese Academy of Sciences

College of Electronic Information and Automation, Civil Aviation University of China

AI问答

Address：No.3888 Dong Nanhu Road, Changchun, Jilin, China Postal code：130033
Tel：0431-86176855 Email：gxjmgc@ciomp.ac.cn
Technical support is provided by Beijing Founder electronics co., LTD 吉ICP备11002662号-17 京公网安备11010802024621
It is recommended to read the content of this site in Chrome&IE9+. Please switch to extreme mode in browser 360.
Cookies We use cookies to help provide and enhance our service and tailor content. By continuing, you agree to the use of cookies.

⁰