CHARACTERIZATION OF ELECTROMAGNETIC BACKSCATTER FROM MOVING TRACKED VEHICLES.

摘要

The characterization of the electromagnetic backscatter from moving tracked vehicles is of significant interest in the area of target classification. The classification process depends upon features in the backscatter signal which are characteristic of different classes of targets.;An electromagnetic backscatter model is developed for the tracks so as to include the effects of the track/ground plane interaction, acceleration, turning, the tracks being out of synchronism and the influence of fenders. The model shows how the motion of the tracks periodically amplitude modulates (PAM) the skin line return thus producing discrete harmonics in the spectrum. These discrete harmonics are shown to undergo many perturbations when the vehicle accelerates and turns.;Three features are then selected based upon the PAM harmonics. These features are extracted from the measured backscatter of tanks, APCs and trucks. Feature plots are made, clustering properties observed and decision regions are then established.;Tables of the probabilities of classification and missclassification for the vehicles are computed and the resulting probabilites indicate that track backscatter will characterize the backscatter from tracked vehicles.;Each radar has a radar cross section determined by the superposition of all of its constituent scattering mechanisms. This cross section will determine the magnitude of the backscattered signal. Additionally, moving subcomponents, such as tracks, will modulate the return signal and these subcomponent modulations can often be used to characterize the target.