Sea clutter model comparison for ship detection using single channel airborne raw SAR data

Abstract

Near real time processing of ship detection in a challenging ocean environment is imperative for maritime security and surveillance. Most of the sea clutter models developed for ship detection have been experimented over fully focussed SAR images. The goal of this paper is to evaluate the performance of clutter models on DLR’s F-SAR acquired singlechannel range-compressed data and range-Doppler image for on-board processing and high resolution ISAR image generation. K-distribution and tri-modal discrete (3MD) texture model were applied over the images and a good fit of these models were observed considering only the main beamclutter of the range-Doppler image. The work was further extended to detect ships at different false alarm rates.