Algorithmic Steps for SAR Backprojection on Radar Based Motion Estimation

When adapting SAR (synthetic aperture radar) techniques to vehicles it becomes obvious, that there are some dramatic differences to air-born SAR systems. The main difference in automotive applications is, that vehicles does not drive with constant speed in straight direction only as planes do. Therefore in this paper a novel algorithm is developed what enables to estimate vehicles ego motion very precisely and at exactly those time steps when it is needed. This algorithm enables both, generating SAR maps as well as autarkic ego motion estimation. Tests in real traffic scenarios show promising results.