An Adaptive Extended Kalman Filter for Target Image Tracking

Abstract

A simple extended Kalman filter has been designed to track targets using outputs from a forward-looking infrared (FLIR) sensor as measurements. It exploits knowledge unused by current correlation trackers - size, shape, and motion characteristics of target, and atmospheric jitter spectral description - to yield enhanced performance. In order to track air-to-air missiles at close range, it incorportateson-line adaptation to target shape effects, changing target motion characteristics, and maximum signal intensity. It is shown to possess considerable performance potential for highly maneuverable targets despite background clutter.