Robust real-time periodic motion detection, analysis, and applications

被引:398
作者
Cutler, R [1 ]
Davis, LS [1 ]
机构
[1] Univ Maryland, Dept Comp Sci, College Pk, MD 20742 USA
关键词
periodic motion; motion segmention; object classification; person detection; motion symmetries; motion-based recognition;
D O I
10.1109/34.868681
中图分类号
TP18 [人工智能理论];
学科分类号
081104 ; 0812 ; 0835 ; 1405 ;
摘要
We describe new techniques to detect and analyze periodic motion as seen from both a static and a moving camera. By tracking objects of interest, we compute an objects self-similarity as it evolves in time. For periodic motion, the self-similarity measure is also periodic and we apply Time-Frequency analysis to detect and characterize the periodic motion. The periodicity is also analyzed robustly using the 2D lattice structures inherent in similarity matrices. A real-time system has been implemented to track and classify objects using periodicity. Examples of object classification (people, running dogs, vehicles), person counting, and nonstationary periodicity are provided.
引用
收藏
页码:781 / 796
页数:16
相关论文
共 36 条
[1]  
ALLMEN M, 1991, THESIS U WISCONSIN M
[2]  
[Anonymous], 1995, VISION INTERFACE
[3]  
[Anonymous], 1970, CLASSICAL DYNAMICS P
[4]  
Brockwell P.J., 1991, TIME SERIES THEORY M
[5]   Recurrence plots revisited [J].
Casdagli, MC .
PHYSICA D, 1997, 108 (1-2) :12-44
[6]  
COHEN CJ, 1996, P IEEE INT C AUT FAC
[7]  
Cutler R., 1999, Proceedings. 1999 IEEE Computer Society Conference on Computer Vision and Pattern Recognition (Cat. No PR00149), P326, DOI 10.1109/CVPR.1999.784652
[8]  
CUTLER R, 1998, P INT C PATT REC AUG, pSA14
[9]   RECURRENCE PLOTS OF DYNAMIC-SYSTEMS [J].
ECKMANN, JP ;
KAMPHORST, SO ;
RUELLE, D .
EUROPHYSICS LETTERS, 1987, 4 (09) :973-977