A Real-time Ship Detetion Method in Sar Images Based on Feature Analysis

This paper focuses on the feature analysis of moving ships in synthetic aperture radar (SAR) images and proposes a detection algorithm for ships. The algorithm first detects all the candidate ships and provides analysis of structural features with the discrete topology. In the analysis of structural features, aspect ratio and ship’s navigation direction can be used to construct feature confidence. Then, the algorithm replaces the pixels’ gray value of the detected ship with the surrounding pixel value. Partial curvilinear structure method is applied on the image, and the gray intensity contrast of the wake to clutter in the image is enhanced. Candidate wakes are extracted from image by using thresholding technique, which can be used for ship discrimination. Finally, the decision making method is applied to discriminate targets based on feature analysis of ship/wake. In this manner, ships on the sea area are detected. Experimental results show that the proposed method can be implemented with time-saving, high-precision ship/wake extraction, feature analysis, and ship recognition.