Title :
Particle Filter Based on Color Feature with Contour Information Adaptively Integrated for Object Tracking
Author :
Pu, Bing ; Zhou, Fugen ; Bai, Xiangzhi
Author_Institution :
Image Process. Center, Beijing Univ. of Aeronaut. & Astronaut., Beijing, China
Abstract :
Particle filter is a probabilistic multi-hypothesis algorithm under the Bayesian framework. In order to establish a robust observing model, in this paper, a novel method which uses a more effective color feature with contour information integrated adaptively is proposed. Experimental results verified that, our approach was efficient.
Keywords :
Bayes methods; image colour analysis; object tracking; particle filtering (numerical methods); probability; Bayesian framework; color feature; contour information integrated adaptively; object tracking; particle filter; probabilistic multihypothesis algorithm; robust observing model; Gray-scale; Histograms; Image color analysis; Lighting; Particle filters; Robustness; Shape; adaptive; color feature; contour information; particle filter;
Conference_Titel :
Computational Intelligence and Design (ISCID), 2011 Fourth International Symposium on
Conference_Location :
Hangzhou
Print_ISBN :
978-1-4577-1085-8
DOI :
10.1109/ISCID.2011.192