Classification of personnel targets by acoustic micro-Doppler signatures

Abstract

Classification of targets using their micro-Doppler signatures has attracted a growing interest in recent years. In addition to their main bulk translation, targets may exhibit additional motions, such as vibrations and rotations, which generate Doppler modulations in the echo that contain unique target features and thus can be used to perform target recognition. Although target classification by micro-Doppler signatures has been exploited in the radio frequency regime for radar systems, much less work has been done in acoustic. In this work, an ultrasound radar operating at 80 kHz has been developed to gather micro-Doppler signatures of personnel targets performing various actions. The performance of a range of classifiers and feature extraction algorithms in distinguishing between these micro-Doppler signatures is presented.