SAR image classification for target detection
Abstract
A simple method for classifying synthetic aperture radar (SAR) images is presented which relies on knowledge of surface roughness and radar backscattering. The proposed simulation algorithm, based on 2-D Fourier transformations, makes it possible to extract various types of urban zones from ground clutter independently on the polarization or inclination. The inverse problem is also treated, and a method for simulating SAR texture on the basis of a 3-D model of surface roughness is proposed. The results obtained are verified on the basis of Varan S images of Neuf-Brisach-Freiburg and Oberpfaffenhofen.
- Publication:
-
In AGARD
- Pub Date:
- September 1991
- Bibcode:
- 1991tcse.agarR....B
- Keywords:
-
- Backscattering;
- Image Analysis;
- Radar Detection;
- Radar Imagery;
- Radar Scattering;
- Radar Targets;
- Synthetic Aperture Radar;
- Target Acquisition;
- Algorithms;
- Clutter;
- Fourier Transformation;
- Surface Roughness;
- Three Dimensional Models;
- Communications and Radar