Exploitation of a priori information for tracking maritime intermittent data sources

This paper presents results achieved by the Authors within the IPP 1 Multi Sensorics program. The addressed research deals with the exploitation of context information, denoted as "Knowledge Base" within the different steps of multi-sensor data processing for surface surveillance. An overview on the research motivations and the application field (e.g., multi-target tracking in the Maritime Scenario) is given with focus on intermittent information sources (i.e., non-continuous sensor reports, due to low-SNR signals, sensor blind areas, communication loss, etc). The Knowledge Base is described by means of mathematical models and the inclusion of the a priori information into the Extended Kalman Filter is presented. Different target motion models are herein considered. Finally, the achievements are summarised in terms of track continuity performance over a live maritime traffic scenario.