Abstract
Synthetic aperture radar (SAR) ship detection has been playing an increasingly essential role in marine monitoring in recent years. The lack of detailed information about ships in wide swath SAR imagery poses difficulty for traditional methods in exploring effective features for ship discrimination. Being capable of feature representation, deep neural networks have achieved dramatic progress in object detection recently. However, most of them suffer from the missing detection of small-sized targets, which means that few of them are able to be employed directly in SAR ship detection tasks. This paper discloses an elaborately designed deep hierarchical network, namely a contextual region-based convolutional neural network with multilayer fusion, for SAR ship detection, which is composed of a region proposal network (RPN) with high network resolution and an object detection network with contextual features. Instead of using low-resolution feature maps from a single layer for proposal generation in a RPN, the proposed method employs an intermediate layer combined with a downscaled shallow layer and an up-sampled deep layer to produce region proposals. In the object detection network, the region proposals are projected onto multiple layers with region of interest (ROI) pooling to extract the corresponding ROI features and contextual features around the ROI. After normalization and rescaling, they are subsequently concatenated into an integrated feature vector for final outputs. The proposed framework fuses the deep semantic and shallow high-resolution features, improving the detection performance for small-sized ships. The additional contextual features provide complementary information for classification and help to rule out false alarms. Experiments based on the Sentinel-1 dataset, which contains twenty-seven SAR images with 7986 labeled ships, verify that the proposed method achieves an excellent performance in SAR ship detection.
Highlights
With the rapid development of spaceborne SAR, such as TerraSAR-X, RADARSAT-2 and Sentinel-1 [1,2,3], synthetic aperture radar (SAR) ship detection has been playing an increasingly essential role in marine monitoring and maritime traffic supervision [4,5,6]
Two experiments are designed to explore the effect of different layer fusions and the influence of outperformance of the proposed method. with other methods indicates the outperformance of the contextual features
With the labeled dataset on Sentinel-1, this paper opens up the possibility of utilizing deep neural neural networks for SAR ship detection
Summary
With the rapid development of spaceborne SAR, such as TerraSAR-X, RADARSAT-2 and Sentinel-1 [1,2,3], synthetic aperture radar (SAR) ship detection has been playing an increasingly essential role in marine monitoring and maritime traffic supervision [4,5,6]. Many investigations relating to ship detection in SAR imagery have been carried out. Traditional methods [7,8,9] detect targets after sea–land segmentation and utilize the hand-crafted features for discrimination, which has poor performance on nearshore areas and has difficulty ruling out false alarms, such as icebergs and small islands. It is necessary to develop detectors with strong feature extraction capabilities to obtain better performances in SAR ship detection. Deep neural networks are capable of feature representation and have been widely applied for object detection [10,11]. They provide a highly promising approach for end-to-end object detection
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
