Title :
Diffusive CFAR & its extension for Doppler and Polarimetric data
Author :
Barbaresco, F. ; Rivereau, N.
Author_Institution :
Thales Air Systems, Surface Radar Business Line, Strategy Technology & Innovation, Limours, France
Abstract :
A new CFAR is introduced based on non-linear PDE (Partial Differential Equation) operators & geometric flows (Laplace-Beltrami & Mean Curvature Flows). This “Diffusive CFAR” smoothes Radar noisy data but preserves discontinuities. This kind of CFAR is well adapted for Radar detection in Littoral Area, where probability of detection in sea/ground clutter transitions and crest line (most threatening areas) could be non optimal by use of classical CFAR. We introduced these new CFAR in the framework of Information Geometry by taking into account radar signal statistics. We propose diffusive CFAR extension for Doppler & Polarimetric data. Finally, real time implementation is described based on “short-kernel” methods.
Keywords :
Anisotropic Diffusion; Chentsov; Information Geometry; Laplace-Beltrami & Mean Curvature Flows; Non-linear CFAR; Siegel metrics;
Conference_Titel :
Radar Systems, 2007 IET International Conference on
Conference_Location :
Edinburgh, UK
Print_ISBN :
978-0-86341-848-8