Multipath Smearing Suppression for Synthetic Aperture Radar Images of Harbor Scenes

Abstract

Due to the periodic and non-periodic variations in the sea surface, smearing is caused by the multiple paths between the sea surface and man-made objects in synthetic aperture radar images of harbor areas. This smearing can cover real targets and lead to false alarms. To derive the relationship between the motion of the sea surface and blurring in synthetic aperture radar images, a sway signal model is established, and the Doppler spectrum of the sea surface is found to undulate for well-focused targets with different shapes. Based on this finding, a subaperture combined detection algorithm based on an inverse coherence factor filter is developed to separate the unwanted pixels from the resultant synthetic aperture radar image. An energy balance is used to suppress interference and maintain the resolution of the real scene. The algorithm can be automatically applied to synthetic aperture radar images. The experimental results with TerraSAR-X spotlight mode data show that this method can effectively detect and mitigate the effects of the time-varying multipath phenomena.