A new approach to moving terrestrial targets recognition using ground surveillance pulse doppler RADARs

In this paper we propose a new automatic target recognition algorithm to recognize and distinguish of three classes of targets: personnel, wheeled vehicles and animals, using a low-resolution ground surveillance pulse Doppler radar. Using the chirplet transformation, a time-frequency signal processing technique, the parameterized radar signal is then used by the Zernike moments (ZM) for the pertinent features of the targets. The current work provides a new approach for multiresolution analysis and classification of non stationary signals with the objective of revealing important features in noisy and cluttered environment. The algorithm is trained and tested on real radar signatures of multiple examples of moving targets from each class. The results shows the proposed algorithm invariancy against speed and orientation of the targets.