A fast algorithm based on two-stage CFAR for detecting ships in SAR images

Ship detection is an important application of SAR imagery in ocean surveillance. After analyzing the statistical characters of sea clutter, a fast algorithm of ship detection in SAR image is proposed in this paper. The method consists of two CFAR detection stages. The first step utilizes a lognormal based CFAR to sort out the potential target pixels at a high false alarm rate; in the second step, these potential targets are refined under a local process of K distribution based adaptive CFAR detection. Space-born SAR images are used to validate this fast detection algorithm, and results show great improvement on efficiency of the proposed method without decreasing detection performance. The fast algorithm satisfies application demands of ship detection in SAR images.