Research on High-Accuracy Data Association of Space-Borne SAR and AIS

Abstract

Many countries attach importance to maritime surveillance for its wide applications in exclusive economic zone surveillance, environmental monitoring and anti-piracy operations, etc. Space-borne Synthetic Aperture Radar (SAR) is optimal for its high resolution over wide swaths and all weather working capabilities for ship surveillance, but limited to the level of SAR imaging and image interpretation. Meanwhile, with rapid development of space-borne Automatic Identification System (AIS), almost real time and global coverage for ship surveillance has become possible, but not all ships operate or equip with AIS terminal transmitter. Space-borne SAR and AIS has cooperative nature, so ship surveillance by integration of them has attracted more attention. This paper focuses on the data association of space-borne SAR and AIS. State-of-the-art association methods have good performance except in the high-sea-state and high-density-shipping situations. To improve the flexibility and validity, we emphasis on analyzing two factors: AIS-projected position and Doppler displacement, which mainly affect the accuracy. Firstly, this paper presents the theory of data association in detail. Then state-of-the-art method for projecting AIS position based on Dead Reckoning is researched on further, and improved method based on Gray Prediction Model MGRM (1, N) is introduced. High-accuracy estimation of Doppler displacement is discussed later profoundly. The application of Point Pattern Matching using Shape Context in the association is also investigated, and the simulation results illustrate that it outperforms current state-of-the-art method both in the precision as well as robustness.