Applying the Hough transform for detecting ground movers in synthetic aperture radar imagery
John Arthur Miller, Robert Linnehan, Armin Doerry · Proceedings of SPIE, the International Society for Optical Engineering/Proceedings of SPIE · 2016
This paper describes the impact of ground target motion in Synthetic Aperture Radar (SAR) and video SAR mode imagery. The observations provide an approach for optimizing algorithms that detect moving targets by using only the magnitude of a SAR image. A slowly moving target at a constant velocity in the along-track direction or accelerating in the cross-track direction often generates a ridge of intensity that is distinguishable from the background clutter. The direction and location of a detected ridge provide information about the motion of the associated target. The ridge can be approximated as a linear feature and detected using the Hough transform. This approach acts as a complement to detecting the radar shadow of a moving target, improving detection probability. The method is robust enough to discriminate between a ridge associated with a moving target and false alarms due to vegetation, boulders, or stationary manmade objects. Simulated and flight test data collected by General Atomics Aeronautical Systems, Inc. (GA-ASI) validate the method.