Abstract:Synthetic Aperture Radar (SAR) systems can be used for discovery and surveillance. According to the current requirements, accurate ship detection is becoming more and more important. However, the interference factors such as coherent speckle noise and strong target sidelobes will reduce the detection accuracy. In this paper, an improved two-parameter Constant False Alarm Rate (2P-CFAR) method is proposed to solve this problem. After using the non-local means (NLM) filtering method to suppress coherent speckle noise, the 2P-CFAR method is performed to obtain candidate targets, and an equal proportion segmentation procedure is carried out to optimize the aspect ratios of the candidate targets and weaken the sidelobes. Experimental results based on the SAR images acquired by the Gaofen-3 satellite verify the performance of the proposed method.

An Improved 2P-CFAR Method for Ship Detection in SAR Images