Spatial Color Image Databases Summarization

Author

Bouguila, N.

Author_Institution

Concordia Inst. for Inf. Syst. Eng., Concordia Univ., Montreal, Que., Canada

Volume

1

fYear

2007

fDate

15-20 April 2007

Abstract

This paper presents a finite discrete mixture model based on both the Dirichlet and the multinomial distributions to add spatial information to color histograms. The estimation of the parameters and the determination of the number of components in our model are based on the classification expectation-maximization approach and the integrated complete likelihood criterion, respectively. The developed model is applied with success for color images databases summarization.

Keywords

expectation-maximisation algorithm; image classification; image colour analysis; statistical distributions; visual databases; Dirichlet distributions; classification expectation-maximization approach; color histograms; finite discrete mixture model; integrated complete likelihood criterion; multinomial distributions; spatial color image databases; Color; Content based retrieval; Data engineering; Histograms; Image databases; Image retrieval; Information retrieval; Information systems; Spatial databases; Systems engineering and theory; Dirichlet distribution; Image databases; finite mixture models; multinomial;

fLanguage

English

Publisher

ieee

Conference_Titel

Acoustics, Speech and Signal Processing, 2007. ICASSP 2007. IEEE International Conference on

Conference_Location

Honolulu, HI

ISSN

1520-6149

Print_ISBN

1-4244-0727-3

Type

conf

DOI

10.1109/ICASSP.2007.366067

Filename

4217239