2009
HMS 2009 Proceeding
Modeling and analysis of marine traffic for anomaly detection (a position paper)
Authors: Wen-Jing Hsu, Hui Fang, and Shell-Ying Huang
Abstract
This paper presents a system for modeling andanalyzing the patterns and anomalies in the vesselmovement from collected sensor data. The major tasksinvolved are (a) Detection of macro-level trafficpatterns, which includes identifying places ofsignificance based on dwelling time or traffic density. Italso includes construction of representative routes ofreal traffic as opposed to the nominal path onnavigational maps. (b) Characterization of vesselmovements: classification of journeys, transitionalprobability between significant places, ship movemodels, interaction model, etc. which describe themoves of the ships; (c) Tracking and prediction ofships, which enables prediction of the future physicalposition of a given ship, and the evolution of multipleinteracting ships. (d) Anomaly detection: Vessels thatbehave anomalously in time, space, or mode areidentified and monitored. We believe that the systemdeveloped will be a useful tool for marine safety andsecurity.
I3M Scientific Sponsors
I3M Industrial Sponsors
I3M Media Sponsors
