This form describes a programmable, autonomous and stand-alone imaging system for the acquisition and processing of images containing subjects whose size is larger than 1cm (e.g. gelatinous zooplankton, fishes, litter, manufacts), form the seafloor or along the water column, in shallow or deep waters. It is capable to recognize and classify the image content through pattern recognition algorithms that combine computer vision and artificial intelligence methodologies. The relevant content of the acquired images is stored on board or transmitted through the communication facilities of the imaging system (e.g. acoustics or satellite modem). The use of such an image acquisition and processing system, installed on fixed or mobile platforms (e.g. cabled or stand-alone observatories, lander, AUV, ARGO Floats, drifter buoys, sea gliders), allows for low cost and long lasting underwater monitoring activities at local or regional scale.
The underwater video systems currently in use for marine monitoring mainly consist of imaging devices installed on fixed platforms wired to a land station (cabled observatories like EMSO or Ocean Network Canada) or operated or deployed by surface vessels, like for example ROV, AUV. The use of wired stations and vessels makes the monitoring activity expensive and necessarily limited in time and space. The proposed intelligent imaging device overcomes all these limits. It has conceived for stand-alone applications, it has a small volume and weight and can be easily installed onboard mobile or fixed platforms. It performs autonomous monitoring activities for long periods (> 24 months), and across wide geographical areas, as it can be easily located and relocated.
Norway, Spain, Italy, France, Germany, Ireland, UK, Netherlands