Tomographic techniques provide 3-D radar signatures
Because two-dimensional (2-D) Fourier transforms require data to be orthogonally aligned and evenly spaced, they must be resampled to a Cartesian grid before they can be processed. This resampling often leads to data loss and loss of resolution in radar images.
Now, Craig Malek (cquest@ c quest.com ) at CompuQuest (Springfield, VA) has developed a technique that uses two-dimensional (2-D) interpolation techniques to eliminate loss of resolution in radar images. And, because the method samples radar in two dimensions, radar images can be reconstructed in three dimensions. To construct three-dimensional (3-D) images, radar data are collected at both surfaces of constant elevation and azimuth. By using bilinear and nearest-neighbor techniques, radar data are then reconstructed in 3-D sample space. Data are then interpolated to Cartesian coordinates using information from neighboring planes.
Malek has tested the technique on a test target of a 1/3 scale model C-29 aircraft with a 17.5-ft wingspan. Although there was loss of detail due to signal averaging, the nose contour is completely averaged in the image because it is shadowed by the rest of the aircraft.