Ship detection based on eigenvalue-eigenvector decomposition and OS-CFAR detector

Maritime surveillance and ship detection are important applications in marine area. Azimuth ambiguities and system noises in SAR image are common phenomena, which may cause false alarms in ship detection. In this paper, a new method based on eigenvalue decomposition and OS-CFAR detector is proposed. The characteristics of ship azimuth ambiguities and systems noises are also analyzed. The method has been experimented in three RADARSAT-2 QUAD Fine mode images and the results show that this method can effectively discriminate ship targets from azimuth ambiguities and system noises. Comparing to single polarization detector, this method has a better performance, especially in images with lots of azimuth ambiguities and system noises.