Intelligent Ocean Governance—Deep Learning-Based Ship Behavior Detection and Application

Abstract

The development and utilization of marine resources by mankind has brought out a series of practical problems such as the destruction of marine ecology, the damage of seabed assets, and the disputes over marine sovereignty. How to use information technology tools to profile and monitor ships, accurately classify and identify ship behaviors through multi-source data fusion analysis, and timely alert and invert abnormal behaviors have become an important means of intelligent ocean governance. In response to the above needs, this paper classifies the ship’s behavior, designs a new data structure ShipInfoSet that represents the ship’s multi-source heterogeneous spatio-temporal information, and proposes a deep learning-based ship behavior-monitoring algorithm ML-Dabs. Accurate identification of the ship’s behavior based on deep learning has realized the monitoring and warning of different types of ships’ profiles and abnormal behaviors. This paper designs an intelligent ocean information port architecture, which can be implemented by deploying the algorithm.