From Cluster Tracking to People Counting

The Cluster Tracker, introduced in previous work, is used to detect, track, split, merge and remove clusters of pixels significantly different from the corresponding pixels in a r eference image. Clusters with common motion are grouped together into super-clusters during off-line processing, and the number of people in each super-cluster is determined by the sizes of the super-clusters and their pattern of merging and splitting. Finally, this information is used to obtain a statistical summary of the behaviour of people in the field of view.

[1]  G. McLachlan,et al.  The EM algorithm and extensions , 1996 .

[2]  Michael Isard,et al.  BraMBLe: a Bayesian multiple-blob tracker , 2001, Proceedings Eighth IEEE International Conference on Computer Vision. ICCV 2001.

[3]  H. Sebastian Seung,et al.  Algorithms for Non-negative Matrix Factorization , 2000, NIPS.

[4]  W. Eric L. Grimson,et al.  Adaptive background mixture models for real-time tracking , 1999, Proceedings. 1999 IEEE Computer Society Conference on Computer Vision and Pattern Recognition (Cat. No PR00149).

[5]  D. Rubin,et al.  Maximum likelihood from incomplete data via the EM - algorithm plus discussions on the paper , 1977 .

[6]  A. Pece.  Tracking of Non-Gaussian Clusters in the PETS2001 Image Sequences , 2001 .

[7]  P. Mahalanobis.  On the generalized distance in statistics , 1936 .

[8]  Dorin Comaniciu,et al.  Real-time tracking of non-rigid objects using mean shift , 2000, Proceedings IEEE Conference on Computer Vision and Pattern Recognition. CVPR 2000 (Cat. No.PR00662).