EU Joint Research Centre

The European Commission's Directorate General (DG) Joint Research Centre supports the DG Fisheries and Maritime Affairs and EU Member States in the execution of the EU Common Fisheries Policy. Among other things, the Common Fisheries Policy sets quotas on what amounts of each fish species EU member states are allowed to catch. As one can imagine, enforcing the policy presents difficulties. The ocean is a big place — how can EU officials know that fishing vessels go where they say they are going, or if non-EU vessels are fishing in EU waters?

JRC researchers have been working towards one possible answer, the "Vessel Detection System" (VDS). VDS is a marriage of billion-dollar space technology, top notch image processing science and standard open source data management tools, which could provide fisheries enforcement officials with near-real-time monitoring capability.

VDS combines information from the Radarsat-1 and ENVISAT synthetic aperture radar (SAR) satellites, merchant marine vessel transponders and fisheries vessel transponders to create an operational picture of known and unknown vessels in European waters. SAR satellites are very good at sensing vessels, because metal ships are very effective radar backscattering targets while water is not.

Figure1 Figure2 Figure3 Figure4

The SAR satellite generates a 300km by 300km swathe over a pre-selected area of interest and the data is beamed to one of three ground receiving stations. The raw data is processed using proprietary SAR software into an image product, which is transferred to the VDS system. The arrival of a new image triggers the VDS which runs a batch process to extract radar signals as geographic points and stores the points in a PostGIS database, along with metadata about the confidence level, size, and original image frame.

At the same time, positions from the merchant marine and fisheries transponder systems are also fed into the PostGIS database via web services and scripts. The transponder information and radar point targets are correlated inside the database to match against known ships with signals and to flag unidentified signals. Finally the data can be sent to enforcement authorities via E-mail and made accessible in a web front-end using UMN Mapserver.

The entire data processing and correlation chain takes about 20 minutes, which means the map information provided to authorities can be suitable for directing surveillance operations. Patrol aircraft in the air can be re-directed to unidentified targets based on the information provided by the VDS.

Guido Lemoine joined the VDS team in 2003, and helped develop the experimental system. "Originally, the group wanted to build the database on Oracle Spatial and the front-end on ArcInfo", said Lemoine, "However, in the end we made a decision to move exclusively to open source components, to avoid lock-in to proprietary solutions." The first software prototype was demonstrated in 2003, and the system has been tested by authorities ever since.

Lemoine's previous open source database experience had been with MySQL, but the lack of spatial functionality in MySQL made PostGIS a better choice. "We have never regretted this choice", says Lemoine.

The PostGIS database plays a central role in the system, as a repository of all the incoming data, from radar and transponders, and the integration location where the data are correlated to track unidentified targets. At any given time, several hundred thousand time-stamped points and polygons are held in the VDS database. In addition, all the base information for map display (bathymetry, shorelines, fisheries regions) are also managed in the PostGIS database.

Successful demonstrations of VDS for monitoring fishing vessels have encouraged the group to pursue further research to support other policy areas, such as "maritime surveillance" duties for safety or security purposes. The group is also experimenting with reconfiguring their near-real-time data processing chain, to reduce the data processing time from sensor to map to just 10 minutes.

For more information:

Joint Research Centre
Fisheries @ the Joint Research Centre

Published October 2006

Do You Have a PostGIS Story?

We are always looking for new stories of people using PostGIS to solve their probems! Drop us a line and tell us your story:

  • What is your organization?
  • What kind of problem did you have?
  • Why did you choose PostGIS?
  • What other options did you have?
  • How did you do your implementation?
  • How is PostGIS working for you now?