Prescreening and discrimation of maritime targets in single-channel SAR images

A targets detection algorithm is proposed for maritime surveillance by single-channel SAR images. It foresees a preliminary prescreening step, carried out using an adaptive threshold algorithm, followed by a discrimination phase, performed by sub-look analysis. The latter discriminates the pixels detected by the former step in three classes, i.e. targets, sea, and azimuth ambiguity. The algorithm is tested on single channel StripMap TerraSAR-X data. Results indicate that the selected algorithm shows promising detection performance and ambiguity rejection capabilities.