Segmentation and classification of combined optical and radar imagery

Author

Schoenmakers, Ronald P H M ; Vuurpijl, Louis G.

Author_Institution

Cortese s.a.s, Ispra, Italy

Volume

3

fYear

34881

fDate

10-14 Jul1995

Firstpage

2151

Abstract

The classification performance of a neural network for combined six-band Landsat-TM and one-band ERS-1/SAR PRI imagery from the same scene is carried out. Different combinations of the data-either raw, segmented or filtered-using the available ground truth polygons, training and test sets are created. The training sets are used for learning while the test sets are used for verification of the neural network. The different combinations are evaluated

Keywords

geophysical signal processing; geophysical techniques; geophysics computing; image classification; image segmentation; neural nets; optical information processing; radar imaging; remote sensing; remote sensing by radar; sensor fusion; spaceborne radar; synthetic aperture radar; ERS-1; IR; Landsat TM; PRI; SAR imaging; geophysical measurement technique; ground truth polygons; image classification; image processing; image segmentation; infrared; land surface terrain mapping; microwave radar; multidimensional signal processing; multispectral remote sensing; neural net; neural network; optical imaging; radar imagery; satellite remote sensing; sensor fusion; six band; synthetic aperture radar; training sets; visible; Image segmentation; Laser radar; Layout; Neural networks; Optical computing; Optical filters; Radar imaging; Remote sensing; Satellites; Testing;

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.1995.524133

Filename

524133