The length estimation of ship targets in ISAR images

Inverse Synthetic Aperture Radar produces images which can be trained to recognize by human operators. Because of varying angular motions (roll, pitch, and yaw), the cross-range resolution is produced. So the Automatic Target Recognition (ATR) of ship can be implemented. The centre-line of ship is an important feature. This paper addresses the problem of centre-line detection and length estimation for ship targets. The aim is to calculate the length of ship for recognition. In order to obtain more accurate results, a new analytical model is proposed, the position of mainmast and the differential Doppler concerning the centre-line is used in the estimation the length of ship. The performance of the proposed algorithm is shown by real data and simulated data.