A 3D TARGET IMAGING ALGORITHM BASED ON TWO-PASS CIRCULAR SAR OBSERVATIONS

Abstract

In circular synthetic aperture radar (CSAR), the radar collects data over a circular not a linear trajectory. The two-dimensional (2D) CSAR image also contains three-dimensional (3D) information about the target. In this paper, we propose an imaging algorithm for 3D target reconstruction with two-pass CSAR observations so as to overcome the problem of limited azimuthal persistence for real anisotropic targets, and avoid the assumption that target falls into the same resolution cell for each elevation pass when multi-pass observations are used. In the algorithm, the first step is to divide both of the two full-aperture CSAR data into subapertures in the same way; the second step is to obtain, for each subaperture, the height of target according to the established relationship between the pixel displacements in the image pair of two observations on the same focal plane and the pixel displacements in the image pair of one observation on two different focal planes; the third step is to obtain the 3D target coordinates based on the retrieved height information and the 2D image coordinates; the last step is to get the final 3D image by combining the obtained 3D images of all subapertures. The results of point target simulation indicate that the 3D information (both amplitudes and positions) are well reconstructed. At the same time, the processing results of backhoe data simulated by the Xpatch software show that the outline of the 3D structure is also well reconstructed although the available data corresponding to the depressing angles are not as good as expected.