Frequency-Diverse Computational Imaging System Enhanced by Synthetic Aperture Radar Techniques

Computational imaging (CI) has been recently proposed as an alternative to conventional synthetic aperture radar (SAR) systems. CI systems rely on antennas radiating a set of spatially-incoherent patterns, enabling a physical layer compression of the information from the scene. This, in turn, simplifies the hardware layer of the imaging system while, at the same time, provides a fast data acquisition speed. This work studies the impact of considering the information from several acquisition positions on the quality of the images retrieved with a CI system. The measurement results obtained with a frequency-diverse CI system show that combining the information from a reduced number of acquisition positions yields a significant improvement on the quality of the retrieved images. In particular, considering approximately two orders of magnitude less samples than those required by a SAR system, high-resolution images with a reduced clutter level can be obtained.