Tracking Multiple Persons under Occlusion Using Appearance Model Based on Attributed Relational Graph

Abstract:

Appearance features are important for tracking persons in stationary scenes. The proposed algorithm was based on appearance mold built by attributed relational graph (ARG). The ARG was used for modeling human appearance features containing color and spatial information. The matching degree of the ARGs was utilized for analyzing tracking situations in current frame. For tracking persons under occlusion, multiple feature patches were generated and the genetic algorithm was used for finding optimal labels of patches. Experiments showed the utility and performance of the proposed approach.

• Periodical: Applied Mechanics and Materials (Volumes 220-223)
• Pages: 1324-1328
• DOI: https://doi.org/10.4028/www.scientific.net/AMM.220-223.1324
• Online since: November 2012
• Authors: Qin Wan, Yao Nan Wang, Hong Shan Yu, Xiao Qi Mo
• Keywords: Appearance Model, Attributed Relational Graph, Occlusion, Tracking Multiple Persons, Visual Surveillance

References

[1] M. Hu, W. Hu, and T. Tan: Proc of ICPR Vol.2 (2004), p.724.

[2] M.B. Capellader, D. Doermann, D. DeMenthon, and R. Chellappa: Proc of ICIP Vol. 5 (2003), p.85.

[3] Q. Zhao and H. Tao: IEEE International Workshop on VS-PETS Vol. 7 (2005), p.263.

[4] Park S, Aggarwal J K,Computer Vision and Image Understanding Vol. 1 (2006), p.1

[5] A. Elgammal, D. Harwood, and L. S. Davis: Proc of ECCV Vol. 3 (2000), p.751.

[6] Yaonan Wang, Qin Wan, Hongshan Yu: J. Sys. Eng. E Vol. 20 (2009), p.1167.

[7] David A. Forsyth and Jean Ponce: Computer Vision: A Modern Approach (Prentice Hall, NY 2003).

[8] W.J. Christmas, J. Kittler: IEEE T. Pattern. Anal Vol. 17(1995), p.749.