Unsupervised change detection techniques using multispectral satellite images

Author

Niemeyer, Irmgard ; Canty, Morton ; Klaus, Dieter

Author_Institution

Res. Centre Julich, Germany

Volume

1

fYear

1999

fDate

1999

Firstpage

327

Abstract

The use of commercial satellite images for the verification of nuclear treaties has been suggested previously in the context of a new protocol extending the rights of the International Atomic Energy Agency (IAEA) to verify the absence of undeclared nuclear activities within its member states. The present paper describes an application of principal component and canonical correlation analysis for change detection of Landsat TM images of the plutonium production complex near Kyshtym, Russia

Keywords

correlation methods; image recognition; principal component analysis; remote sensing; unsupervised learning; IAEA; International Atomic Energy Agency; Kyshtym; Landsat TM images; Russia; canonical correlation analysis; commercial satellite images; multispectral satellite images; nonproliferation treaties; nuclear treaties; plutonium production complex; principal component analysis; undeclared nuclear activities; unsupervised change detection techniques; Covariance matrix; Eigenvalues and eigenfunctions; Geography; Group technology; Image analysis; Principal component analysis; Protocols; Remote sensing; Satellites; Vectors;

fLanguage

English

Publisher

ieee

Conference_Titel

Geoscience and Remote Sensing Symposium, 1999. IGARSS '99 Proceedings. IEEE 1999 International

Conference_Location

Hamburg

Print_ISBN

0-7803-5207-6

Type

conf

DOI

10.1109/IGARSS.1999.773487

Filename

773487