Marine radar properties, analysis and applications

Abstract

In this thesis marine radars are compared with synthetic aperture radars (SAR) and the possibility of cross-over applications are investigated. A first cross-over has been demonstrated by using the TS-CFAR on marine radar images. The TS-CFAR was originally developed for SAR and is a constant false alarm rate (CFAR) detection algorithm based on truncated statistics. Detecting weak targets embedded in sea clutter is difficult because it is hard to find a model describing the sea in its various conditions. For detection algorithms such as constant false alarm rate (CFAR) the accuracy of the sea clutter estimation directly relates to performance. A model comparison between Weibull-, Gamma-, Log-normal- and Rayleigh PDF on sea clutter data has been performed. Sea clutter data was separated at sectors at different range, and both low- and high resolution data from a X-band Sea-Hawk radar and S-band JRC JMR solid state radar was used in the comparison. Our results show that the Weibull PDF models the sea clutter best in most cases, but we found indications that the Gamma PDF fit the clutter best when texture is present. The TS-CFAR was compared to a CA-CFAR both with an underlying Weibull model and an underlying Gamma model. Our results show that TS-CFAR clearly performed best on high resolution data and for multiple target situations. This thesis also shows how to extract raw data from marine radars which proved to be a formidable task.