Classification of dual-polarisation meteorological radar data

A preliminary study of data processing algorithms, oriented toward the classification of meteorological echoes from a linear dual-polarization radar, is presented. The main aim is to detect and recognize ground clutter even whenever it is superimposed on meteorological echoes. A further objective is the discrimination between echoes from rain and solid-phase precipitations. Two different methods for texture-based analysis, following both statistical and structural approaches, are proposed. Parameters extracted are fed to a maximum-likelihood classifier to discriminate the different classes of interest. The proposed algorithms have been tested on a set of data acquired by the CP-2/RP6 radar during the survey campaign Maypole-84. Experimental results are reported and briefly discussed.