Feature extraction for radar target recognition using time sequences of radar cross section measurements

The radar cross section (RCS) signatures are not only related to the scattering structure of targets, but also related to the modulation of target micro-motion. In this paper, a feature extraction procedure for radar target recognition based on Mellin transform of the time sequences of RCS measurements is proposed. The mathematical relationship between target scattering distribution in the cross range and the RCS amplitude is derived and analyzed. The RCS features of targets are extracted using sequential method. Finally the feasibility of target recognition using Mellin transform and the accuracy of the algorithm are shown by simulated data sets.