计算机科学 ›› 2014, Vol. 41 ›› Issue (10): 306-309.doi: 10.11896/j.issn.1002-137X.2014.10.064

• 图形图像与模式识别 • 上一篇    下一篇

基于双矩形框标定的改进TLD算法

张伟伟,汤光明,孙怡峰   

  1. 信息工程大学 郑州450001;信息工程大学 郑州450001;信息工程大学 郑州450001
  • 出版日期:2018-11-14 发布日期:2018-11-14
  • 基金资助:
    本文受河南省科技攻关基金(122102210047)资助

Improved TLD Algorithm Based on Region Marking of Double Bounding Boxes

ZHANG Wei-wei,TANG Guang-ming and SUN Yi-feng   

  • Online:2018-11-14 Published:2018-11-14

摘要: 鉴于基于单个矩形框标定的Tracking-Learning-Detection (TLD) 算法无法兼顾跟踪目标的“重点性”和“完整性”,提出了一种基于双矩形框标定的改进算法。在标定整个目标区域的矩形框的基础上,算法在目标变化相对稳定的区域标定另一个矩形框,以指示跟踪的重点区域。在提取跟踪点时,采用分配权重的方法使重点区域产生更多的跟踪点,从而提高算法对局部变化的适应能力。实验表明,改进后的算法在跟踪局部保持稳定而其余部分有所变化的目标上有较高的性能提升;而对于不存在稳定局部区域的目标,跟踪效果没有明显改善。

关键词: TLD,区域标定,物体跟踪,矩形框

Abstract: Considering that the Tracking-Learning-Detection (TLD) framework can hardly balance the tracking of the whole and the interest,by using region-marking method of single bounding box,this paper proposed an improved algorithm introducing a region-marking method of double bounding boxes.With one bounding box labeling the whole object,the algorithm draws another bounding box over the stable area of image to indicate the interested part.While extracting the trace points,the weighting approach is adopted to produce more points in interested area,which improves TLD’s adaptability to local variance.Experimental results show that the improved algorithm has good performance in tracking the object when a fixed part keeps stable but the rest varies.For the object not containing the stable local area,the effects are not so obvious.

Key words: TLD,Region marking,Object tracking,Bounding box

[1] Wang Qing,Chen Feng,Xu Wen-li,et al.An experimental com-parison of online object-tracking algorithms[C]∥Proceedings of SPIE:Image and Signal Processing.2011:138-147
[2] Zdenek K,Jiri M,Krystian M.Tracking-learning-detection[J].IE-EE Trans on Pattern Analysis and Machine Intelligence,2012,4(7):1409-1422
[3] Zdenek K,Jiri M,Krystian M.Forward-Backward Error:Automatic Detection of Tracking Failures[C]∥International Conference on Pattern Recognition (ICPR).2010:23-26
[4] Zdenek K,Jiri M,Krystian M.P-N Learning:Bootstrapping Binary Classifiers by Structural Constrains[C]∥Computer Vision and Pattern Recognition (CVPR).2010:49-56
[5] 程立英,张丹,赵姝颖,等.一种基于TLD改进的视觉跟踪算法[J].科学技术与工程,2013,3(9):2382-2385
[6] Lucas B D,Kanade T.An iterative image registration technique with an application to stereo vision[C]∥Proceedings of Imaging Understanding Workshop.1981:121-130
[7] 朱克忠.基于光流法对移动目标的视频检测与应用研究[D].合肥:合肥工业大学,2007
[8] Tomas V.Long-term Visual Object Tracking with online Lear-ning [D].Czech Technical University,2012
[9] Wang K,Babenko B,Belongie S.End-to-End Scene Text Recognition[C]∥2011 IEEE International Conference on Computer Vision (ICCV).2011:1457-1464
[10] Rupali S C,Patil.Object Tracking Based on Tracking-Learning-Detection . http://gnebehay.github.io/OpenTLD/genbehay_thesis_msc.pdf
[11] Li Tian-yu,Liu Jia,Gong chen,et al.Robust Object Tracking by Combining Boosting Learning and Particle Filter[J].Journal of Computational Information Systems,2012,2(8):9593-9601
[12] Zadenek k,Krystian M,Jiri M.Face-TLD:Tracking-learning-detection applied to faces[C]∥ICIP.2010:26-29
[13] The standard TLDdata.http://personal.ee.surrey.ac.uk/Personal/Z.Kal-al/index.html

No related articles found!
Viewed
Full text


Abstract

Cited

  Shared   
  Discussed   
No Suggested Reading articles found!