Optimizing Vessel Trajectory Compression

Thanks to the Automatic Identification System (AIS), tracking vessels across the seas provides a powerful means for maritime safety and environmental protection. However, large amounts of streaming AIS positional updates from vessels can hardly contribute additional knowledge about their actual motion patterns. Vessels are generally expected to maintain straight, predictable routes at open sea, except in cases of adverse weather conditions, accidents, traffic restrictions, etc. In [11] a maritime surveillance system was introduced, involving a trajectory detection module that can provide summarized representations of vessel trajectories by consuming AIS positional messages online. The key idea behind the proposed summarization is that keeping only some critical points may be enough to reconstruct with tolerable accuracy the original course of each vessel. Indeed, instead of retaining every incoming position for every vessel or even applying a costly multi-pass trajectory simplification algorithm, this method drops positions along trajectory segments of "normal" motion characteristics. In addition, the retained critical points can be marked with suitable annotations, i.e., indicating stops, turning points, changes in speed, etc. The resulting trajectory synopsis per vessel is derived from those judiciously annotated critical points and can approximately reconstruct its original course.