Recognition of Military and Civilian ships in SAR Images based on Ellipse Fitting Similarity

Abstract

With the development of SAR imaging technology, detailed information of ship targets is displayed in high resolution SAR, which facilitates target recognition. In this paper, the differences between the structures of military and civilian ships are investigated. The paper utilizes the characteristics of military ships that have structures like towering bridges, which make them have a protruding structure in SAR images compared with civilian ships. As a result, the similarity between the ship contour and the fitting ellipse is used to recognition the military and civilian ships. Hence, a method for recognizing military and civilian ships in high-resolution SAR images based on ellipse fitting is proposed. This method can be divided into two steps, first extracting the edge of the ship target in the image, and then judging the similarity between the edge and the ellipse. The high-resolution images obtained by GF-3 satellite are used for experimental verification. The experimental results show that this method can effectively classify military and civilian ships.