Online Mobility Tracking Against Evolving Maritime Trajectories

We examine techniques concerning mobility tracking over trajectories of vessels monitored over a large maritime area. We focus particularly on maintaining summarized representations of such trajectories in online fashion based on surveillance data streams of positions relayed from a fleet of numerous vessels using the Automatic Identification System. First, we review generic, state-of-the-art simplification algorithms that can offer concise summaries of each trajectory as it evolves. Instead of retaining every incoming position, such methods drop any predictable positions along trajectory segments of “normal” motion characteristics with minimal loss in accuracy. We then discuss online filters that can reduce much of the noise inherent in the reported vessel positions. Furthermore, we present a method for deriving trajectory synopses designed specifically for the maritime domain. With suitable parametrization, this technique incrementally annotates streaming positions that convey salient trajectory events (stop, change in speed or heading, slow motion, etc.) detected when the motion pattern of a given vessel changes significantly. Finally, we discuss a qualitative comparison of maritime-specific synopses along with trajectory approximations obtained from generic simplification algorithms and highlight their pros and cons in terms of approximation error and compression ratio.