A Fast Generation Algorithm of Radar Images for Ships on Time-evolving Sea Surface

This paper presents a fast simulation algorithm for radar images of ship on dynamic sea surface. The instantaneous radar observation angles are first derived from the 6-degrees of freedom (DOF) pose of ship driven by sea waves. Taking a specific observation angle and radar parameters as input, on the one hand, we can predict the two dimensional (2-D) complex scattering image of ship via high-frequency electromagnetic (EM) calculation and projection technique, efficiently. On the other hand, the 2-D complex scattering function of the rough sea surface can be calculated by using the sea spectrum inversion and capillary wave modification facet scattering model (CWMFSM), directly. Finally, a radar image for the composite scene can be obtained by superimposing the complex scattering images of target and rough sea surface, after which a fast point-spread operation using 2-D window function is made in the frequency domain for sidelobe suppression. The proposed approach can generate single radar image of ship on sea surface in a few seconds compared with traditional algorithms always including time-consuming processing steps. So, this algorithm can be utilized to simulate the snapshot image sequence of ship on time-evolving sea surface under the given radar observation trajectory, which provides abundant pre-training samples for target recognition. Numerical results demonstrate the effectiveness of the proposed technique.