Extraction and analysis of structural features of ships in high resolution SAR images

Ship classification in high resolution SAR satellite imagery is a hotspot and preceding problem in SAR applications. According to the different features of cargo ship, oil tanker and container ship in high resolution SAR images, we propose a method to extract structure feature of three civilian ship types. Firstly, the morphological pre-processing is used for sidelobe suppression. Secondly, the main axes are extracted using Hough Transform; the Minimum Enclosing Rectangle(MER) of ship is obtained and rotated to the direction of principal axes, then it is adjusted to fit the ship well. Finally, the scattering properties of different ship types are analyzed. The result shows that this method has good performance in feature extraction of ships in high resolution SAR imagery.