• INITIAL PARAMETERS ANALYSIS FOR MIXTURE OF GAUSSIAN MODEL
Abstract
The mixture of Gaussian is a common model for background subtraction. There are several parameters in such a model. Obviously, the assignment of initial values to these parameters affects the accuracy of background subtraction. In this paper, we analyze in detail the impact of different initial parameter values based on the EM algorithm. The tested results of waving trees video sequences have illustrated. This parameter values analysis provides suggestions how to choose suitable initial parameter values while using a mixture of Gaussian model in video surveillance application
Keywords
Full Text:
PDFRefbacks
- There are currently no refbacks.
This work is licensed under a Creative Commons Attribution-NonCommercial-ShareAlike 4.0 International License.
© 2011-2024 Research Journal of Pure Algebra (RJPA) Copyright Agreement & Authorship Responsibility | Counter |