Airfield Runway Detection from Synthetic Aperture Radar Image

An algorithm of airfield runway detection from high resolution synthetic aperture radar image is proposed. Runways are identified by using Hough transform technique, followed by a series of speckle suppression and region of interest (ROI) segmentation. First, speckle suppression based on edge preserving is described. Then, constant false alarm rate detector with probability iteration is used for ROI segmentation. Finally, peak values detected from Hough transform domain determine the edge location of runway. High resolution images are used to illustrate the algorithm, and experimental results show that performance is satisfactory.