Linear filtering of spatially invariant image sequences for feature separation under three types of image noise

Author

Olmstead, Reed ; Farison, James B.

Author_Institution

Dept. of Eng., Baylor Univ., Waco, TX, USA

fYear

2002

fDate

2002

Firstpage

162

Lastpage

166

Abstract

Many important imaging applications in medical imaging, remote sensing and other areas result in a set of images of the same scene with no relative scene-sensor motion and in which the pixel intensities are the linear sum of the contributions of the distinct features in the scene. Such image sets are called linearly additive, spatially invariant (LA SI) image sequences. Previous research has shown, both mathematically and with examples, that a K-image sequence with M distinct features (M\n\n\t\t

Keywords

feature extraction; filtering theory; image sequences; noise; Gaussian noise; K-image sequence; LA SI image sequences; Poisson noise; feature separation; image noise; linear filtering; linearly-additive spatially-invariant image sequences; medical imaging; pixel intensities; remote sensing; salt-and-pepper noise; spatially invariant image sequences; Additive noise; Biomedical engineering; Biomedical imaging; Feature extraction; Gaussian noise; Image sequences; Layout; Maximum likelihood detection; Pixel; Remote sensing;

fLanguage

English

Publisher

ieee

Conference_Titel

System Theory, 2002. Proceedings of the Thirty-Fourth Southeastern Symposium on

ISSN

0094-2898

Print_ISBN

0-7803-7339-1

Type

conf

DOI

10.1109/SSST.2002.1027026

Filename

1027026