Ship detection with TanDEM-X data: a statistical test for a polarimetric notch filter

Ship detection is an important topic for security and surveillance of maritime and coastal areas. A solution exploiting satellite SAR sensors is particularly interesting. SAR polarimetry has a valuable contribution in improving ship detection capabilities. In this paper a statistical test is devised for a polarimetric ship detector, namely the Geometrical Perturbation-Polarimetric Notch Filter. The detector was showing good performance compared with other detectors however, the selection of the threshold was following a physical (or asymptotic) rationale. Here a study on the statistical distribution of the detector output is performed in order to design a Constant False Alarm Rate detector. The algorithm is tested on TanDEM-X data acquired in the North Sea under severe weather conditions. The data are also accompanied by acquisitions of AIS data.