eISSN:2278-5299

International Journal of Latest Research in Science and Technology

DOI:10.29111/ijlrst   ISRA Impact Factor:3.35

A News Letter Sign UP!
ENFORCED INDIVIDUALIZATION OF HUMAN CLUSTER IN TRACKING BY MIXTURE OF GAUSSIAN

Research Paper Open Access

International Journal of Latest Research in Science and Technology Vol.3 Issue 2, pp 197-200,Year 2014

ENFORCED INDIVIDUALIZATION OF HUMAN CLUSTER IN TRACKING BY MIXTURE OF GAUSSIAN

Youngsung Soh, Yongsuk Hae, Hadi Ashraf, Mudasar Qadir, Intaek Kim

Correspondence should be addressed to :

Received : 24 April 2014; Accepted : 27 April 2014 ; Published : 30 April 2014

Share
Download 125
View 178
Article No. 10303
Abstract

Vision based surveillance system plays an important role in security field. Automatic tracking of objects such as human and vehicle is vital in such a system. In human tracking, it frequently happens that moving people get merged and demerged. When merging occurs, the tracking system should know what and how many humans are in the cluster. When cluster gets demerged into individuals, the tracking system should be able to differentiate among them to cope with correct human following. It will be useful in tracking if we can enforce the human cluster, where multiple humans adhere, to be broken into individual human. In this paper, we propose a novel method for individualization of human cluster in tracking. The method consists of three steps. They are edge finding and edge projection, mixture of Gaussian fitting, and cutting point decision. We applied the method to various scenarios of human cluster and obtained good results.

Key Words   
Gaussian, human tracking, cluster tracking, tracking
Copyright
References
  1. Sharma,” A Blob Representation for Tracking Robust to Merging and Fragmentation,” Proc. Of IEEE Workshop on Application of Computer Vision, pp.161-168, 2012.
  2. Kumar, S. Ranganath, K. Sengupta, and H. Weimin,“Cooperative Multitarget Tracking With Efficient Split and Merge Handling,” IEEE Trans. on Circuits and Systems for Video Technology,vol.12, pp. 1477–1490, 2006.
  3. Genovesio and J. C. Olivo-Marin,”Split and Merge Data Association Filter for Dense Multitarget Tracking,” Proc. Of 17th International Conf. on Pattern Recognition, vol. 4, pp.677-680. , 2004.
  4. J Garcia, J. M. Molina, J. A. Besada, and J. I. Portillo,”A Multitarget Tracking Video System based on Fuzzy and Neuro-fuzzy techniques,”Journal of Applied Signal Processing. Special Issue on Advances in Intelligent Vision Systems, vol.14, pp.2341-2358, 2005.
  5. Stauffer and W. Grimson, "Adaptive background mixture models for real-time tracking," IEEE International Conference on Computer Vision and Pattern Recognition, pp.246- 252, 1999.
  6. Power and J. Schooness, "Understanding background mixture models for foreground segmentation," Image and Vision Computing New Zealand, pp.267-271, 2002.
To cite this article

Youngsung Soh, Yongsuk Hae, Hadi Ashraf, Mudasar Qadir, Intaek Kim , " Enforced Individualization Of Human Cluster In Tracking By Mixture Of Gaussian ", International Journal of Latest Research in Science and Technology . Vol. 3, Issue 2, pp 197-200 , 2014


Responsive image

MNK Publication was founded in 2012 to upholder revolutionary ideas that would advance the research and practice of business and management. Today, we comply with to advance fresh thinking in latest scientific fields where we think we can make a real difference and growth now also including medical and social care, education,management and engineering.

Responsive image

We offers several opportunities for partnership and tie-up with individual, corporate and organizational level. We are working on the open access platform. Editors, authors, readers, librarians and conference organizer can work together. We are giving open opportunities to all. Our team is always willing to work and collaborate to promote open access publication.

Responsive image

Our Journals provide one of the strongest International open access platform for research communities. Our conference proceeding services provide conference organizers a privileged platform for publishing extended conference papers as journal publications. It is deliberated to disseminate scientific research and to establish long term International collaborations and partnerships with academic communities and conference organizers.