计算机科学 ›› 2017, Vol. 44 ›› Issue (3): 278-282.doi: 10.11896/j.issn.1002-137X.2017.03.057

• 图形图像与模式识别 • 上一篇    下一篇

一种基于Mean Shift的快速跟踪算法

邹青志,黄山   

  1. 四川大学电气信息学院 成都610065,四川大学计算机学院 成都610065
  • 出版日期:2018-11-13 发布日期:2018-11-13

Fast Tracking Algorithm Based on Mean Shift Algorithm

ZOU Qing-zhi and HUANG Shan   

  • Online:2018-11-13 Published:2018-11-13

摘要: 针对Mean Shift算法难以跟踪快速运动目标、算法迭代次数多以及耗费时间长的问题,提出了一种基于Mean Shift的快速运动目标检测方法,该方法结合帧差法并融合背景信息来快速检测运动目标;同时提出一种新的相似性度量方法进行初步检测,排除干扰并快速选出符合标准的目标以进行Mean Shift匹配,找出最佳目标。该方法不仅减少了传统方法的迭代次数,缩短了算法所需时间,而且在跟踪实验中取得了较好的跟踪效果,提升了算法的鲁棒性。

关键词: Mean Shift算法,帧差法,目标跟踪,快速跟踪,鲁棒性

Abstract: A fast moving target detection method based on Mean Shift was proposed for the problems that the Mean Shift algorithm is difficult to track fast moving objects,the number of iterations of the algorithm is too large and the process is time consuming.The method is combined with frame difference method and fuses background information for rapid detection of moving target.A new similarity measure method for preliminary testing was put forward to exclude the interference and fast select targets in accordance with the standard Mean Shift matching,finding out the best target.This method not only reduces the number of iterations of the traditional method,but also reduces the time required for the algorithm,and it achieves better tracking performance in the tracking experiment,which improves the robustness of the algorithm.

Key words: Mean Shift algorithm,Frame difference method,Target tracking,Fast tracking,Robustness

[1] LI X H,ZHANG T Y,SHEN X D,et al.Object tracking using an adaptive Kalman filter combined with mean shift [J].SPIE Journal of Optical Engineering,2010,9(2):020503-020503-3.
[2] YAN J,WU M Y,CHEN S Z,et al.Anti-occlusion tracking algorithm based on Mean Shift and fragments [J].Optics and Precision Engineering,2010,18(6):1413-1419.(in Chinese) 颜佳,吴敏渊,陈淑珍,等.跟踪窗口自适应的Mean Shift跟踪[J].光学精密工程,2010,8(6):1413-1419.
[3] SHEN C H,BROOKS M J,VAN DEN HENGEL A.Fast Glo-bal Kernel Density Mode Seeking:Applications to Localization and Tracking [J].IEEE Transactions on Image Processing,2007,6(5):1457-1469
[4] MAGGIO E,CAVALLARO A.Multi-part target representation for color tracking [C]∥IEEE International Conference on Ima-ge Processing.2005:729-732.
[5] LIANGFU L,ZUREN F,et al.A coarse-to-fine kernel matching approach for mean shift based visual tracking [J].Opto-Electronics Review,2007,7(1):65-71.
[6] WANG T,LIU W N,HAN W L,et al.Target tracking algo-rithm based on improved mean shift[J].Chinese Journal of Li-quid Crystals and Displays,2012,7(3):396-400.(in Chinese) 王田,刘伟宁,韩伟良,等.基于改进Mean Shift的目标跟踪算法[J].液晶与显示,2012,7(3):396-400.
[7] COLLINS R T,YANX L,LEORDEANU M.Online selection of discriminative tracking features[J].IEEE Trans on Pattern Analysis and Machine Intelligence,2005,7(10):1631-1643.
[8] ADAM A,RIVLIN E,SHIMSHON L.Robust fragments-based tracking using the integral histogram[C]∥IEEE Computer Society Conference on Computer Vision and Pattern Recognition.2006:798-805.
[9] HAO X B.Background modeling and dynamic partition basedobject tracking [J].Electronics R&D,2010,7(10):21-23.(in Chinese) 韩晓波.基于背景建模和动态分块的目标跟踪[J].电子技术研发,2010,37(10):21-23.
[10] LI S X,CHANG H X,ZHU C F.Adaptive pyramid mean shift for global real-time visual tracking [J].Image and Vision Computing,2010,8(3):424-437.

No related articles found!
Viewed
Full text


Abstract

Cited

  Shared   
  Discussed   
No Suggested Reading articles found!