Computing with words in maritime piracy and attack detection systems

Abstract

In this paper, we propose to apply recent advances in deep learning to design and train algorithms to localize, identify, and track small maritime objects under varying conditions (e.g., a snowstorm, high glare, night), and in computing-with-words to identify threatening activities where lack of training data precludes the use of deep learning. The recent rise of maritime piracy and attacks on transportation ships has cost the global economy several billion dollars. To counter the threat, researchers have proposed agent-driven modeling to capture the dynamics of the maritime transportation system, and to score the potential of a range of piracy countermeasures. Combining information from onboard sensors and cameras with intelligence from external sources for early piracy threat detection has shown promising results but lacks real-time updates for situational context. Such systems can benefit from early warnings, such as “a boat is approaching the ship and accelerating,” “a boat is circling the ship,” or “two boats are diverging close to the ship.” Existing onboard cameras capture these activities, but there are no automated processing procedures of this type of patterns to inform the early warning system. Visual data feed is used by crew only after they have been alerted of a possible attack. Camera sensors are inexpensive but transforming the incoming video data streams into actionable items still requires expensive human processing.