A Maritime Situational Awareness Framework Using Dynamic 3D Reconstruction in Real-Time

Abstract

Accessible evaluation of the security status of maritime infrastructures is a key factor for maritime safety. Conveying security-related information at a glance can support the decision-making process of stakeholders, such as port authorities, law enforcement agencies and emergency services. In this work, we present a novel maritime situational awareness framework that visualises dynamic 3D point clouds and static 3D geospatial data in a web-based display. Our system can capture, reconstruct and display dynamic maritime objects in real-time in 3D allowing the user to interactively assess the situation around a maritime infrastructure. To allow scaling and remote deployment we use a 3D reconstruction system that runs on an embedded system. Resulting point clouds are streamed to a server that further processes the data to generate semantic information in context with precalculated geospatial data. Using a web-based viewer we present dynamic point clouds and static geospatial data in 3D as well as additional semantic metrics to users. We verify our system using synthetic photo-realistic images from a virtual camera. The system updates in near real-time (5 Hz) providing users with a full 3D maritime situational awareness map.