A robust variation of the principle components algorithm

Author

Haberstroh, Richard ; Madonna, Richard

Author_Institution

Res. & Dev. Center, Northrop Grumman Corp., Bethpage, NY, USA

Volume

2

fYear

34881

fDate

10-14 Jul1995

Firstpage

1269

Abstract

Discusses two remote sensing image classifiers for hyperspectral data that are relatively insensitive to small errors in the atmospheric transmission function. These classifiers permit the authors to use laboratory measured spectral databases for classification of unknown spectra. Numerical results are presented that demonstrate that the classifiers have a better than 90% identification accuracy even when using the “wrong” atmospheric transmission function

Keywords

geophysical signal processing; geophysical techniques; image classification; optical information processing; remote sensing; atmospheric transmission function; geophysical measurement technique; hyperspectral; image classification; land surface; multispectral remote sensing; optical imaging; principle components algorithm; robust variation; terrain mapping; visible infrared IR; Atmospheric measurements; Atmospheric modeling; Building materials; Classification algorithms; Hyperspectral imaging; Hyperspectral sensors; Image databases; Laboratories; Layout; Meteorology; Remote sensing; Robustness;

fLanguage

English

Publisher

ieee

Conference_Titel

Geoscience and Remote Sensing Symposium, 1995. IGARSS '95. 'Quantitative Remote Sensing for Science and Applications', International

Conference_Location

Firenze

Print_ISBN

0-7803-2567-2

Type

conf

DOI

10.1109/IGARSS.1995.521722

Filename

521722