Comparison and visualization of feature space behaviour of statistical and neural classifiers of satellite imagery

Author

Fierens, F. ; Kanellopoulos, I. ; Wilkinson, G.G. ; Mégier, J.

Author_Institution

Joint Res. Centre, Inst. for Remote Sensing Applications, Varese, Italy

Volume

4

fYear

1994

fDate

8-12 Aug 1994

Firstpage

1880

Abstract

Currently both statistical and neural classifiers are being used for the classification of multispectral satellite imagery. Because both classifier types are being used as `black boxes´ and because they are values based on different mathematical models the reasons for their different performance levels are not well understood. The authors have used visualization of class decision boundaries in feature space as a means to gain insight into the classification processes

Keywords

geophysical signal processing; geophysical techniques; image classification; image colour analysis; infrared imaging; neural nets; optical information processing; remote sensing; IR imaging; class decision boundaries; feature extraction; feature space behaviour; geophysical measurement technique; image classification; land surface terrain mapping; multispectral method; neural net; optical imaging; remote sensing; satellite imagery; statistical classifier; visible; visualization; Data mining; Data visualization; Electronic mail; Expert systems; Mathematical model; Neural networks; Remote sensing; Satellites; Testing; Three dimensional displays;

fLanguage

English

Publisher

ieee

Conference_Titel

Geoscience and Remote Sensing Symposium, 1994. IGARSS '94. Surface and Atmospheric Remote Sensing: Technologies, Data Analysis and Interpretation., International

Conference_Location

Pasadena, CA

Print_ISBN

0-7803-1497-2

Type

conf

DOI

10.1109/IGARSS.1994.399600

Filename

399600