Surveillance of the Greek Sea Border via the Synergistic Use of Satellite and Terrestrial Technologies
Gardikis, Georgios; Costicoglou, Socrates; Andrikopoulos, Ilias

Border Security is a topic of considerable interest at EU level. Especially for Greece, Sea Border Security is of profound importance; Greece has a coastline of ~17,300 km, the longest of any other EU/Med country. Almost half of this coastline corresponds to the continental part of the country and a little more than half to the archipelagic complexes of the Aegean and Ionian Seas (about 3,000 islands, a few hundreds of which are inhabited - more than 9,800 if we also count the rocky formations). Especially when considering illegal immigration, Greece exhibits one major characteristic not encountered in other EU countries: apart from the illegal immigration from north of Africa, a major part of this illegal activity is localized on the eastern Aegean sea covering a wide geographical area which presents very short distances (in some cases <10 nm) from Turkey.

Earth Observation (EO) techniques for vessel detection on SAR and optical images have been proved to be a valuable tool in maritime surveillance, exhibiting significant accuracy in detecting vessels in open sea over a wide coverage area. However, due to the increased response time and the inability to detect very small vessels, EO alone is often not sufficient to address use case scenarios within the Aegean Sea, where illegal migration routes can be very short. In order to effectively address this challenge, satellite observation tools must be promptly complemented with surveillance mechanisms based on terrestrial data, such as optical and thermal imaging cameras, radars, fixed ground sensors, manned patrolling aircrafts, Vessel Monitoring Systems (VMS), Automatic Identification Systems (AIS), even Unmanned Aerial Vehicles (UAV). While ground systems can provide very short response times, their main limitation is their limited coverage.

In this context, the paper discusses a promising approach towards an improved border security of extended coverage, which is also investigated at EU and National (Greek) level in response to various policies and regulations. This approach involves the integration and correlation of information derived from EO satellites with in-situ sensor data towards the production of an Integrated Maritime Picture (IMP). The paper presents the long-term evolution of an experimental/pre-operational sea border surveillance system especially tailored for the Greek scenario. This evolution has taken place over the last few years and has been carried out in the frame of several projects, such as the ESA STINGRAY, ESA MARISS and MARISS Scale-Up, FP7 DOLPHIN and the recently commenced JASON project, funded by the Greek Secretariat of Research and Technology. All of these projects build on satellite Earth-Observation and also Satellite Communications platforms to build an integrated multi-sensor system for sea border surveillance.

The paper intends to provide a technical insight into the advancements carried out within each of these projects. The following aspects (here being mentioned only in brief) are analysed and discussed:

  • EO SAR data, accompanied with vessel detection reports, is correlated to AIS (Automatic Identification System) reports in order to identify legal vessels among the ones which are detected in the SAR image. The data fusion process includes algorithms for both spatial and temporal alignment.
  • The maritime picture is further enhanced by combining EO and AIS information with in-situ sensor data. Coastal infrared (thermal) cameras and coastal radars are used to quickly detect small targets at distances relatively close to the project. Object recognition mechanisms are employed to isolate the vessels in the captured image, while depth extraction techniques from disparity maps are used to derive the vessel position; again, in addition to camera/radar calibration, temporal/spatial alignment is carried out so as to correctly project the targets to the EO satellite image;
  • low-cost Unmanned Aerial Vehicles (UAVs) are also exploited, which examine a specific target at a pre-defined location. The series of photographs which are captured are analysed on the ground in near-real time via advanced photogrammetry algorithms, also achieving target identification via feature extraction and analysis. The identified targets are projected on the EO image, along with their recorded routes.

    All the aforementioned technical activities are being carried out in close cooperation with the respective national authorities (Hellenic Navy and Hellenic Coast Guard), who provide not only guidelines with respect to service requirements/areas-of-interest but also useful feedback on the outcome of the service. The aim is to deploy the composite service into fully operational phase within the next few years.