Affiliation:
1. Tianjin University of Technology
Abstract
Target tracking algorithm is used widely in many fields, such as robot vision system, intelligent surveillance and medicine, but computational complexity and lack of dedicated embedded system for real-time processing have affected its application. This paper presents a method that combines embedded system, smart camera and mobile robot for detecting and tracking the moving targets. On the basis of particle filter algorithm, mean shift embedded particle filter algorithm is proposed and implemented on embedded platform with ARM+DSP dual core framework. At last, the whole system is optimized to improve the real-time property. The proposed method has a very powerful data processing ability, which can offer a high reliability for the navigation of a mobile robot.
Publisher
Trans Tech Publications, Ltd.
Reference7 articles.
1. K. Quast and A. Kaup: Scale and shape adaptive mean shift object tracking in video sequences, Proceedings of 17th European Signal Processing Conference (EUSIPCO), (2009), p.1513–1517.
2. D. Comaniciu and P. Meer: Mean Shift analysis and application, Proceedings of the Seventh IEEE International Conference on Computer Vision, vol. 2 (1999), p.1197–1203.
3. M. Melzi, A. Ouldali and Z. Messaoudi: Multiple Target Tracking Using The Extended Kalman Particle PHD Filter, Proceedings of EUSIPCO 2010, Aalborg, Denmark, 2010, p.1824–1825.
4. J.H. Kotecha and P.M. Djuric: Gaussian Particle Filtering, IEEE Trans. on Signal Process, Vol. 51(10) (2003), pp.2592-2601.
5. U.C. Jung, H.J. Seung, et al.: Multiple objects tracking circuit using Particle Filters with multiple features. Proceedings of 2007 IEEE International Conference on Robotics and Automation, 2007, pp.4639-4644.