A SHIP GHOST INTERFERENCE REMOVAL METHOD BASED ON GAOFEN-3 POLARIMETRIC SAR DATA

During Synthetic Aperture Radar (SAR) imaging, the presence of ghost is frequently observed on SAR images of maritime scenes due to the finite pulse repetition frequency and non-ideal antenna pattern. In Polarimetric Synthetic Aperture Radar (PolSAR) images of ships, the ship movement makes the dispersion of span which is called ghost interference. This problem leads to high false alarm rates and measurement errors. To solve this problem, we propose a method applied to full-polarimetric SAR data. Firstly, we use multi-feature combination to enhance the scattering mechanism of the targets. Secondly, based on this method, using the Rank-1 and generalized similarity parameter (GSP) to improve the contrast between the sea and the ships. Finally, interference feature filter (IFF) is used to get the image with interference removed. We use the GaoFen-3 (GF-3) full-polarimetric SAR data for experiments. The results show that this method effectively removes the ghost interference, and then we will perform a target detection to prove that it can reduce the false alarm rate.