A Ship Wake Detection Method in SAR Images Based on Joint Sparse Representation

The strong sea clutter and noise in synthetic aperture radar (SAR) images will cover weak ship wakes, making it difficult for conventional ship wake detection methods to accurately detect ship wakes. A ship wake detection method based on joint sparse representation is proposed in this paper to improve the detection effect of ship wakes in complex backgrounds. Firstly, the sparse representation model of SAR images based on the Radon transform is constructed. Secondly, using dual-domain images in the spatial and gradient domains, the optimization function for ship wake detection based on joint sparse representation is constructed and solved iteratively. Then, the weighted clustering criterion based on sparse coefficient and Radon coordinates is used to perform sparse coefficient clustering, and the centerline detection of the ship wake is realized by the inverse Radon transformation of the cluster center. Finally, the experimental results of TerraSAR-X images are used to verify the effectiveness of the proposed method.