Ghost Removal of Compact Polarimetric SAR Ships Based on Multifeature Collaboration and Enhancement

Compact Polarimetric (CP) SAR has a unique advantage in marine target observation, which can obtain rich polarization information and maintain a large observation width. However, due to the Doppler effect, moving ship targets are prone to ghost during the imaging process. Since CP data lose some scattering information, it is more difficult to distinguish the target from interference. This paper proposes a multi-feature collaboration and enhancement method for removing ghosts from the CP SAR ship target. The ship and ghost interference features are enhanced by multi-feature collaborations, then the contrast between ship and ghost interference is enhanced using an extremum separation feature descriptor, and finally, the ghost interference is removed using a signal-to-clutter ratio based Interference Feature Filter (IFF). Given limited CP data for ships with ghost phenomenon, the experiments simulate Circular Transmit and Linear Receive (CTLR) mode data using GF-3 fully polarimetric data, and the results show that this method can remove the ghost around the ship in the CP image better.