The detection of small fishing ships is very important for maritime fishery supervision. However, it is difficult to detect small ships using synthetic aperture radar (SAR), due to the weak target scattering and very small number of pixels. Polarimetric synthetic aperture radar (PolSAR) has been widely used in maritime ship detection due to its abundant target scattering information. In the present paper, a new ship detector, named ΛM, is developed based on the analysis of polarization scattering differences between ship and sea, then combined with the two-parameter constant false alarm rate method (TP-CFAR) algorithm to conduct ship detection. The goals of the detector construction are to fully consider the ship’s depolarization effect, and further amplify it through sliding window processing. First, the signal-to-clutter ratio (SCR) enhancement performance of ΛM for ships with different lengths ranging from 8 to 230 m under 90 different combinations of windows are analyzed in detail using three set of RADARSAT-2 quad-polarization data, then the appropriate window size is determined. In addition, the SCR enlargement between ΛM and some typical polarization features is compared. Among these, for ships of length greater than 35 m, the average contrast of ΛM is 33.7 dB, which is 20 dB greater than that of the HV channel. For small vessels of length less than 16 m, the average contrast of ΛM is 16 dB higher than that of HV channel on average. Finally, the RADARSAT-2 data including nonmetallic small vessels are used to perform ship detection tests, and the detection ability for conventional and small ships of some classic algorithms are compared and analyzed. For large vessels of length greater than 35 m, the method proposed in this paper is able to obtain a superior detection result, maintain the ship contour well, and suppress false alarms caused by the cross side lobe in the SAR image. For small vessels of length less than 16 m, the method proposed in this paper can reduce the number of missed targets, while also obtaining superior detection results, especially for small nonmetallic vessels.

No abstract

As artificial intelligence continues to advance, deep learning has greatly contributed to the advancement of ship recognition using Synthetic Aperture Radar (SAR) images. Deep learning-based SAR ship recognition performance is largely dependent on the sample set used. SAR ship recognition datasets published in recent years, however, are most derived from a single SAR satellite sensor. It needs to be evaluated and analyzed carefully whether the model trained by a single satellite dataset can still achieve the same accuracy with different SAR satellites. The paper focuses on the following research to address these issues. Firstly, using multiple SAR satellite sensors, we create a new SAR ship dataset (named generalization performance evaluation dataset, GPED) containing multi-resolution and multipolarization data to examine the generalization performance of the deep learning-based SAR ship recognition method. GPED and a marine target detection dataset (MTDD) are then used to evaluate and analyze the generalization performance of current mainstream deep learning methods. According to the the experiment results, the mean average accuracy (mAP) of the ship recognition model trained on GPED is generally higher than that of MTDD, which proves that GPED has a better generalization performance. Furthermore, SAR ship detection datasets have more samples than ship recognition datasets, which inspired us to use transfer learning to transfer knowledge from ship detection to ship recognition. In this paper, a method for ship recognition based on transfer learning that utilizes the knowledge gained from the ship detection task is proposed. The method includes two modules: pretraining module and fine-tuning module. It can apply samples of unlabeled ship types to ship recognition, thus reducing the number of labeled samples that are required for ship recognition. The experimental results on GPED and MTDD show that our method can achieve good recognition performances.