A new approach to coherent change detection in VideoSAR imagery using stack averaged coherence

Abstract

Coherent Change Detection (CCD) is the title given to a process used to detect differences between pairs of Synthetic Aperture Radar (SAR) images. In this process, the estimated coherence between a pair of complex SAR images is calculated and used to infer the presence or absence of any changes. Typically, if the estimated coherence is high there is no change whereas low estimated coherence provides an indication of change. VideoSAR is a land imaging mode where the radar is operated in the spotlight mode for an extended period of time. A sequence of imagery is continuously formed via back-projection to a common Cartesian grid while the radar platform is either flying by or circling the target. Whereas CCD was originally designed to look for differences between a single pair of SAR images, in this paper CCD is extended to VideoSAR imagery demonstrating the benefits of averaging coherence across a “stack” of images. The paper reviews the DRDC X-band Wideband Experimental Airborne Radar, describes the CCD workstation developed for conventional SAR and VideoSAR imaging, reviews the CCD algorithms, describes a CCD experiment and presents some experimental results.