An approach based on mean shift and KALMAN filter for target tracking under occlusion

This paper combines the mean shift algorithm with the Kalman filer for target tracking. First, the starting position of mean shift is found by the Kalman filter, then the mean shift uses it to track the object position. The occlusion problem is a difficult problem during target tracking. When severe...

Full description

Saved in:
Bibliographic Details
Published in2009 International Conference on Machine Learning and Cybernetics Vol. 4; pp. 2058 - 2062
Main Authors Jie Zhao, Wen Qiao, Guo-Zun Men
Format Conference Proceeding
LanguageEnglish
Published IEEE 01.07.2009
Subjects
Cybernetics
Economic forecasting
Educational institutions
Electronic mail
Feature extraction
Histograms
Kalman filter
Kalman filters
Kernel
Machine learning
Mean shift
Occlusion
Target tracking
Online AccessGet full text

Cover

More Information
Summary:This paper combines the mean shift algorithm with the Kalman filer for target tracking. First, the starting position of mean shift is found by the Kalman filter, then the mean shift uses it to track the object position. The occlusion problem is a difficult problem during target tracking. When severe occlusion problem takes place, a novel method is proposed to solve this problem in this paper. In that case, the predictive position of the Kalman filter is regarded as its measured value. Make the Kalman filter has the ability to estimate the coming state. Then using the mean shift algorithm find the accurate target position in current frame. Experimental results show that the proposed algorithm is very effective to solve the occlusion problem.
ISBN:9781424437023
1424437024
ISSN:2160-133X
DOI:10.1109/ICMLC.2009.5212129