Ship detection with adaptive parameter based on detection background analysis

Contaminated with various clutter, such as sea clutter, ionospheric interference, earth echoes, in HF surface wave radar (HFSWR), detection background seldom follows a uniform probability distribution. The ideal suppositions by the classical detection algorithms limit the detection performance of the operational system. Thus this paper proposes a new detection architecture which can classify and identify the detection background in real time to improve the detection performance. Firstly, the detection background is classified into five parts with a criterion based on the results of background region segmentation. Then, using the statistical analysis of the classified detection background, a detection method with adaptive parameters is proposed which can select the proper background distribution parameters based on the type of the reference cells. Finally, real data are used for verifying the advantages of the proposed method comparing with three detection algorithms for HF radar introduced in literature recently.