Advanced Ship Detection for Spaceborne based Maritime Awareness
Radius, Andrea1; Ferreira, Joćo1; Carmo, Paulo1; Marques, Paulo2

In the last years the increase in marine traffic generated the necessity of global monitoring for marine environment management in terms of safety, security and fisheries.
The increasing number of new satellite-based Synthetic Aperture Radar (SAR) systems, and the intrinsic capability of the transmitted electromagnetic pulses to interact with the ships and to retrieve its cinematic characteristics, made this instrument particularly fit to improve global maritime awareness through the fusion with other cooperative data (AIS, VMS, LRIT).

The growing need of global maritime awareness gave a push to the realization of different projects in the European context, each one focused on a different particular objective.
Particularly useful is the synergy between the operational and research aspects, being the goal of the last to improve the state of the art in the field of ship detection. Two European projects are the key to strive this synergy: the project MARitime Security Service (MARISS), which implements the operational capability, and the R&D Dolphin project, which is focused on the deep exploitation of remote sensing data and on the technological development of advanced techniques for ship detection and classification purposes.

Particularly, the key maritime surveillance requirements that should be obtained in the context of the aforementioned projects are:

  • the capability to detect small and fast boats;
  • the capability to track the targets of interest in open seas;
  • the capability to characterize the detected targets in terms of geographical position, size, velocity and ship class;
  • the capability to discriminate potential illegal activities through anomalous behavior analysis;

    This paper introduces the multipurpose Edisoft Vessel Detection software (EdiVDC) implemented by the EDISOFT company, which comes from the necessity to respect increasingly stringent requirements in terms of ship detection.

    The EdiVDC software is being operationally used in the framework of the MARISS project and it integrates advanced processing algorithms developed in the scope of the Dolphin project with the cooperation of ISEL-IT (Instituto de Telecomunicaces), improving the software capability and introducing new functionalities.

    Presently, the software includes the capability of detection and classification from different types of SAR data, which allows the characterization of the detected targets in terms of time, position, size and velocity. It also provides a correlation functionality which allows correlating non cooperative detections (made through SAR imagery) with the available cooperative information (AIS, VMS, LRIT) for detection of suspect vessels.
    Finally, the software will allow the "Local Area Analysis" functionality, with the aim to perform a dedicated analysis in a local area of interest, to find potential moving and small targets - the results of this analysis shall largely depend on the resolution, data type and will be constrained by the physical limitations. In this work we present the functionalities of the software and the main results achieved with different types of SAR data.