Coherent texture decomposition using AM-FM model

Author

Nguyen, Chuong T. ; Havlicek, Joseph P.

Author_Institution

Sch. of Electr. & Comput. Eng., Univ. of Oklahoma, Norman, OK, USA

fYear

2010

fDate

23-25 May 2010

Firstpage

81

Lastpage

84

Abstract

We introduce a novel decomposition algorithm capable of extracting locally coherent and visually meaningful texture components from images. The algorithm estimates texture dominant orientation for each coherent component and iteratively extracts it from the image based on a new quantitative coherency measure formulated in the modulation domain. The original image is perfectly reconstructed from extracted components.

Keywords

amplitude modulation; feature extraction; frequency modulation; image reconstruction; image texture; AM-FM model; coherent component; decomposition algorithm; image reconstruction; modulation domain; quantitative coherency measure; texture dominant orientation; texture extraction; visually meaningful texture components; Amplitude modulation; Filter bank; Frequency modulation; Humans; Image decomposition; Image reconstruction; Iterative algorithms; Signal processing; Signal resolution; Speech analysis; AM-FM image model; image decomposition;

fLanguage

English

Publisher

ieee

Conference_Titel

Image Analysis & Interpretation (SSIAI), 2010 IEEE Southwest Symposium on

Conference_Location

Austin, TX

Print_ISBN

978-1-4244-7801-9

Type

conf

DOI

10.1109/SSIAI.2010.5483912

Filename

5483912