A Finely Focusing Method of SAR Using Very Deep Neural Network

Doppler-parameter estimation is a basic auto-focusing method of SAR. However, as the stability of the platform decreases, the Doppler-parameter method cannot achieve high estimation accuracy, and thus leads to the degradation of image quality. In order to solve the problems, this paper proposed a model of SAR signal finely focusing based on Very Deep Neural Network. After roughly focused, SAR signal is sent to pre-trained network to get finely focused result. Experiments on simulated data and actual data shows that the image quality has been greatly improved, meanwhile, the PSLR and ISLR of the result are -14.5494dB and -13.8190dB, which indicates that this method is suitable for solving SAR signal finely focusing problem.