SIR-C polarimetric image segmentation by neural network

Author

Sergi, R. ; Satalino, G. ; Solaiman, B. ; Pasquariello, G.

Author_Institution

Dipartimento di Fisica, GNCB-CNR, Bari, Italy

Volume

3

fYear

1996

fDate

27-31 May 1996

Firstpage

1562

Abstract

In this paper, the results of the segmentation process of polarimetric multiband SAR images are shown. Purpose of the work is the image interpretation in absence of ground-truth. The segmentation process is performed by the self organizing map network which is an unsupervised neural network. The objective of the segmentation is the selection of homogeneous regions on the image and the results are evaluated in terms of grey level statistics on same restricted areas (urban and salina areas)

Keywords

geophysical signal processing; image segmentation; radar imaging; radar polarimetry; self-organising feature maps; spaceborne radar; statistical analysis; synthetic aperture radar; unsupervised learning; SIR-C polarimetric image segmentation; grey level statistics; homogeneous regions; image interpretation; multiband SAR images; salina areas; self organizing map network; unsupervised neural network; urban areas; Covariance matrix; Image resolution; Image segmentation; Neural networks; Organizing; Phase measurement; Radar polarimetry; Scattering; Statistics; Telecommunications;

fLanguage

English

Publisher

ieee

Conference_Titel

Geoscience and Remote Sensing Symposium, 1996. IGARSS '96. 'Remote Sensing for a Sustainable Future.', International

Conference_Location

Lincoln, NE

Print_ISBN

0-7803-3068-4

Type

conf

DOI

10.1109/IGARSS.1996.516731

Filename

516731