Hard and Soft Data Fusion for Maritime Traffic Monitoring Using the Integrated Ornstein-Uhlenbeck Process

When commonly used Automated Identification System (AIS) data to monitor vessel traffic is unavailable, one can resort to other types of data such as remote sensing and open source data to improve detection and tracking in order to increase Maritime Domain Awareness (MDA). The paper proposes a method for fusion of soft open source data (e.g. Twitter, Farkwar and other), remote sensing (RADARSAT-2) and S/T-AIS (hard) data in application to maritime traffic monitoring in the coastal and the open waters. The method assumes modeling vessels' motion by the Integrated Ornstein-Uhlenbeck process. The data fusion performance evaluation using real data is presented showing improved tracking for MDA especially in the absence of AIS data. The impact of inclusion of soft, in particular the Twitter data, into the fusion module is also discussed.