|
Abstract:
|
Synthetic aperture radar (SAR ) is an imaging technique based on the radio reflectivity of the target being imaged . SAR instruments offer many advantages over optical imaging due to the ability to form coherent images in inclement weather , at night , and through ground cover . High resolution is achieved in azimuth through a synthesized aperture much larger than the physical antenna of the imaging device . Consequently , proper focusing requires accurate information about the relative motion between the antenna phase center and the scene . Any unknown target velocity , acceleration , rotation , or vibration will introduce errors in the image . This work addresses a novel method of focusing a moving target in a SAR image through the estimation of various motion parameters . The target azimuth position is determined through monopulse radar , at which point range velocity and acceleration are estimated across a series of overlapping sub -apertures . Cross -range velocity is then estimated through a search to optimize an image quality metric such as entropy or contrast . A final focused image is then generated based on this velocity vector . Methods of extending this work for a single phase center system are considered . This technique is demonstrated with real radar data from an experimental system , and the performance of this technique is compared both subjectively and with a variety of image metrics to the MITRE keystone technique . Finally , extensions to this current line of research are considered . |