Share Email Print

Proceedings Paper

An improved Gaussian mixture model
Author(s): Dayong Gong; Zhihua Wang
Format Member Price Non-Member Price
PDF $14.40 $18.00
cover GOOD NEWS! Your organization subscribes to the SPIE Digital Library. You may be able to download this paper for free. Check Access

Paper Abstract

An improved Gaussian mixture model is presented to substitute the typical method of Chris Stauffer which revealed its weakness in uncontrollability of the background constructing course and foreground mergence time as well as invalidation to the low duty background. By setting appropriate time parameters which meet the monitoring needs, the improved method effectively controls the estimates updating process of each background in Gaussian mixture model via layered attenuating the estimates and intensifying the recurrence events while requires almost the same computation. The simulation of traffic monitoring videos indicates that: this model has no scraps of provisionally staying objects, efficaciously picks up the low duty background.

Paper Details

Date Published: 14 March 2013
PDF: 9 pages
Proc. SPIE 8768, International Conference on Graphic and Image Processing (ICGIP 2012), 87682C (14 March 2013); doi: 10.1117/12.2010876
Show Author Affiliations
Dayong Gong, Hangzhou Polytechnic (China)
Zhihua Wang, Chongqing Univ. of Technology (China)

Published in SPIE Proceedings Vol. 8768:
International Conference on Graphic and Image Processing (ICGIP 2012)
Zeng Zhu, Editor(s)

© SPIE. Terms of Use
Back to Top