Multispectral correlation moving target detection and tracking

Abstract

This paper presents feasibility studies of a multispectral moving target detection (MTD) system designed to detect and track low-signature moving targets. The system performs multispectral detection based on linear and nonlinear maximum discrimination and representation features (MRDF) for low-signature target detection and correlation tracking of slowly moving targets in clutter. Multispectral detection was tested against multispectral data from an acousto-optic tunable filter (AOTF) in the visible and near infrared. The feasibility of correlation target tracking was demonstrated with actual Army MICOM FLIR long wavelength infrared (LWIR: 8 to 12 rim) images ofmoving vehicles and tanks in clutter.