In recent years, European maritime countries had to face new situations involving the traffic of illegal vessels. The fight against unauthorized fishing, irregular migration and related smuggling activity has become an international priority.
Image processing and computer vision methods can contribute to find a solution to these problems by using the large scale data which is nowadays made available by satellite constellations.
With this aim, researchers at the Signals and Images Lab of ISTI-CNR have proposed a system to automatically detect and recognize all the vessels within a given area. The maritime satellite imagery is automatically processed to extract visual informative features of the candidate vessels and to assign an identification label to each candidate. Besides the analysis of the vessel itself, also the associated wake generated by the ship's motion, is considered for detailed analysis, since it represents an important source of information. Indeed, the spectral content of the wave components in the wake, are related to the velocity of the ship itself through known hydrodynamics relationships. Hence, provided the image resolution is large enough to observe the individual wake features, a possible approach to estimate the ship's kinematics is to perform a frequency analysis on the wake area to detect the wavelength related to the ship velocity.
The features extracted through satellite image analysis are then exploited to perform a vessel classification, implemented by a classifier which has been trained using an existing knowledge base. The algorithm performs successfully in 75% of cases, featuring an almost linear time complexity with respect to the input data.
Such research is being carried out within the European Space Agency project "Optical/SAR data and system Integration for Rush Identification of Ship models" (OSIRIS) which has been launched in March 2016, with the primary purpose of developing a software platform dedicated to maritime surveillance.
For more information, please visit: http://si.isti.cnr.it/index.php/hid-project-category-list/44-project-osiris-page