An automatic radar based aerial target recognition framework

Abstract

Helicopters and Air Launched Cruise Missiles (ALCM) usually move at constant speeds as compared with other aerial targets. The distinguishing features for helicopters are the rotatory motions of its main and tail rotors whereas for ALCM are their rotational motion and wings. The present study proposes an automatic target recognition system framework for recognition of aerial targets, specifically Helicopters and ALCM. The study generates simulations of Micro-Doppler signatures for helicopters and ALCMs, followed by identification of their unique features for their classification. The study further applies various classification techniques on simulated dataset so produced to identify the aerial target. The classification results so produced by various classifiers are analyzed and compared, the best results to identify the Helicopters and ALCM, using their Micro-Doppler signatures, are produced by Support Vector Machine classifier.