Multiple Objects Tracking Based on Mixture Features

Abstract:

For multiple objects tracking in complex scenes, this paper proposes a new tracking algorithm for multiple moving objects. The algorithm makes likelihood calculation by using new DG_CENTRIST feature and color feature, and then calculates the overlapping ratio of the tracking object and the object in the current frame using coincidence degree to measure the occlusion. The algorithm has good robustness and stability, and the experiment results show that this method can effectively improve the accuracy of the multiple target tracking.

• Periodical: Advanced Materials Research (Volumes 945-949)
• Pages: 1869-1874
• DOI: https://doi.org/10.4028/www.scientific.net/AMR.945-949.1869
• Online since: June 2014
• Authors: Dong Mei Li*, Tao Li
• Keywords: CENTRIST Feature, Object Tracking, Particle Filter

References

[1] Y. Rui, Y. Chen. Better Proposal Distributions: Object Tracking Using Unscented Particle Filter. IEEE Conference on Computer Vision and Pattern Recognition (2011), pp.786-793.

DOI: 10.1109/cvpr.2001.991045

[2] K. Okuma, A. Taleghani, N. De. Freitas: A boosted particle filter: multitarget detection and tracking. Proc of the European Conference on Computer Vision (2004), pp.28-39.

DOI: 10.1007/978-3-540-24670-1_3

[3] P. Viola, M. Jones: Rapid Object Detection using a Boosted Cascade of Simple Features. IEEE Conference on Computer Vision and Pattern Recognition (2001).

DOI: 10.1109/cvpr.2001.990517

[4] J. Vermaak, A. Doucet, P. Perez: Maintaining Multi-Modality through Mixture Tracking. International Conference on Computer Vision (2003).

DOI: 10.1109/iccv.2003.1238473

[5] J. Wu, Geyer, C. Rehe: Real-time human detection using contour cues. IEEE International Conference on Robotics and Automation (2011), pp.860-867.

DOI: 10.1109/icra.2011.5980437