A ship detection and tracking method with time sequential shipborne radar imagery

Abstract

ABSTRACT Ship detection and tracking is significant to the safe navigation, harbor surveillance and naval defence. In practice, the ship cannot be properly detected and tracked due to the influence of the ambient or its moving attitude on its echo in some images of the sequence. To solve this problem, we proposed a method for ship detection and tracking with time sequential shipborne radar imagery. The method consists of the Kalman filter and the correlation matching technique. The Kalman filter is used to predict the search areas of the ship in the next frame. The correlation matching technique is applied to extract the position of the ship, which can automatically adjust the parameters of the Kalman filter. If the tracked ship is incapable of detection in current frame, to keep the ship tracking successively, the predicted position obtained by applying the Kalman filter can be used as the current position of the ship. It has been tested with 46 temporally consecutive X-band shipborne radar images to demonstrate the validity of the method. The result shows that the ship in the shipborne radar images can be properly de tected and tracked. Keywords: Ship tracking, ship detection, shipborne radar, Kalman Filter