VICEN AI-based automatic detection system

The analysis of sea clutter done in the work on artificial intelligence for automatic vessel detection helped CMRE to find a new model to estimate winds from radar images.

Raúl Vicen, a scientist leading the "Decisions in Uncertain Ocean Environments" project at CMRE, has been working at the Spanish University of Alcalá (UAH) for several years on his PhD thesis devoted to "Automatic Detection of Signals by Using Artificial Intelligence Techniques".

Recently, his effort has been recognized by the Spanish Royal School of Telecommunication Engineers (COIT) and the Spanish Association of Telecommunication Engineers (AEIT) as the best Spanish PhD thesis on Defence and Security. "I am really proud of it – he says – that's great to see that, when you are strongly committed to do something, you can obtain really exciting results".

The thesis presents a new system based on artificial intelligence (AI) capable of automatically detect vessels of different shapes and sizes at sea under different meteorological and oceanographic (METOC) conditions. The system stores up knowledge about the maritime environment acquired during a learning stage and is able to generalize this knowledge to other environments never seen during its learning stage. Its ability to learn from the environment and to constantly adapt its behaviour reduces false alarms.

This result has been considered particularly significant for defence and security applications, where the mitigation of the consequences of additive interference in signal detection and recognition stages is critical. The knowledge acquired by Dr Vicen in his PhD thesis on sea clutter analysis led him to work at CMRE on retrieving wind information from the sea surface clutter observed by marine radars. This work was published in the Journal of Atmospheric and Oceanic Technology. This fact increases the number of parameters a radar can estimate, i.e. not only dedicating a radar to detect, track and recognize vessels, as well as to predict their positions and movements, but also to make METOC measurements.