A generic Bayesian Network for identification and assessment of objects in maritime surveillance

Identification and assessment of objects are key capabilities of surveillance and information systems for maritime environments. Bayesian methodology provides established instruments for fusion of uncertain identification and assessment indications from various information sources. Particularly, Bayesian Networks are well qualified for modeling and inference in this application context. In order to standardize and simplify generation of Bayesian Identification Networks for new operational scenarios, a generic Bayesian Network for identification is proposed. This generic model defines adequate and necessary node types, well suited for modeling of identification and assessment tasks. In addition, the model provides a dependency structure by subgraphs, which simplifies provision of more complex cause-effect relations in Bayesian Identification Networks. Nevertheless, the generic model is flexible enough to cover various application scenarios with manifold operational demands. Closing, generation of a Bayesian smuggler detection Network for maritime surveillance in an exemplary littoral application scenario is presented.