Wideband Tomographic Super-Resolution Radar Image

Abstract

Three-dimensional radar tomography exploits geometric and spatial diversity to reconstruct three-dimensional images. As the bandwidth is independent from the resolution of the tomographic radar imagery, we are proposing a novel method to increase the resolution with more detected features at less computational cost. In this method, we detect a plurality of frequencies of a target and generate a low-resolution image for each frequency with a first-order Born approximation, maintaining the linearity of the inverse problem. Additionally, a high-frequency low-resolution image will be selected as a reference image from the plurality of low-resolution images. Further, we obtain the pixel values of the reference image and compare them to each of the low-resolution images. The method will also include determining new features in each low-resolution image based on the comparisons, generating a high-resolution grid based on the new features identified by the comparisons via interpolation, performing deconvolution of the high-resolution grid using a Wiener filter, and outputting a high-resolution image.