Synthetic aperture radar image processing using cellular neural networks

Author

Kent, Sedef ; Ucan, Osman Nuri ; Ensari, Tolga

Author_Institution

Dept. of Electr. & Electron. Eng., Istanbul Tech. Univ., Turkey

fYear

2003

fDate

20-22 Nov. 2003

Firstpage

308

Lastpage

310

Abstract

In this paper, Cellular Neural Networks (CNNs) have been applied to noisy Synthetic Aperture Radar (SAR) image to improve its performance and appearance. The image has been obtained from Erzurum, Turkey. Because of the importance of imaging quality and appearance for remote sensing applications, CNN has been applied to data for image processing applications that for noise filtering and edge detection. In training, Recurrent Perceptron Learning Algorithm (RPLA) is used as a learning algorithm. According to templates SAR-image has been tested and obtained satisfactory results.

Keywords

geophysical techniques; geophysics computing; neural nets; radar imaging; remote sensing; remote sensing by radar; synthetic aperture radar; CNN; Erzurum; RPLA; Recurrent Perceptron Learning Algorithm; SAR; Turkey; cellular neural networks; edge detection; noise filtering; noisy synthetic aperture radar; remote sensing; Azimuth; Cellular neural networks; Energy resolution; Image processing; Laser radar; Optical imaging; Pulse measurements; Radar antennas; Radar imaging; Synthetic aperture radar;

fLanguage

English

Publisher

ieee

Conference_Titel

Recent Advances in Space Technologies, 2003. RAST '03. International Conference on. Proceedings of

Conference_Location

Istanbul, Turkey

Print_ISBN

0-7803-8142-4

Type

conf

DOI

10.1109/RAST.2003.1303925

Filename

1303925