Adaption of Fast Factorized Back-Projection to Automotive SAR Applications

This paper presents an approach to create a high-resolution image from the surrounding area of a car by utilizing the synthetic aperture radar (SAR) concept for automotive applications. We investigate SAR image formation algorithms which lead to high image quality with acceptable computational complexity. Therefore, we modify the fast factorized back-projection (FFBP) algorithm to make it suitable for ground vehicle applications. Initially, this method was considered for platforms with high range and low azimuth variation geometry. The restrictions of the original approach with regard to our data acquisition scenario are overcome by introducing an additional processing step. The proposed approach has been verified through simulation and experimental data with a 77-GHz radar system in various automotive measurement settings.