A technology allowing real-time tracking and characterization of objects on moving floors
The research conducted by Dr. Frédéric BOUCHARA’s team within the « Signal-Images » (SIIM) activity of the Laboratoire d’Informatique et Systèmes (LIS UMR 7020 Aix-Marseille University, University of Toulon, CNRS) has led to a set of results responding to different applications for real-time monitoring and characterization of objects of interest on moving funds. These results concern a set of software, algorithms and know-how that meet the major challenges of maritime surveillance.
The team has designed a complex video processing function that can be integrated into a global surveillance system whose objective is to detect, locate and characterize objects of interest evolving in the monitored area.
One of the building blocks of this innovative technology is an image processing algorithm that automatically extracts the contour of a rigid object (ship) on a moving floor (water) from a short video sequence. The method is very robust to acquisition conditions (presence of foam or wake, specular reflection, low image contrast).
It is of great interest to the industrial world, particularly in maritime surveillance (combating asymmetric threats, port surveillance, and so on) and related fields. One of the objectives of this innovation was to develop a system that allows an operator to determine by designation (pointing) the angular coordinates of a manoeuvring threat evolving on a water body and to transmit them to a weapon system.
This functional brick also makes it possible to precisely identify a type of ship by deep learning.
The technologies from the laboratory are available for licensing.
Project financed with the support of the European Union with the European Regional Development Fund