Politecnico di Torino - Corso Duca degli Abruzzi, 24 - 10129 Torino, ITALY

+39 011 090 6100 info@tech-share.it

Deep-Eye: acquisition and processing of underwater images

automatic image processingAutonomous Underwater VehicleESTECHsatellite data transmissionstand-alone underwater deviceunderwater image recognition


Deep-Eye: Stand-alone imaging device for the underwater monitoring, for time periods extended in time, conceived to be installed on fixed or mobile platforms (e.g. observatories, AUVs). It consists of an image acquisition sensor, an image processing unit capable to run image recognition algorithms and a data communication system for transmitting the information extracted from the acquired images.

Technical features

The anthropic impacts and the rapid consequences of climate change affect the health of the marine environment. The traditional image-based observation systems make the marine monitoring activity expensive and limited in space (wired stations) and over time (boats). Deep-Eye thanks to its high autonomy and its small size can be easily positioned and repositioned within the area of interest, operating in large areas and for long periods of observation (over 12 months). Deep-Eye integrates hardware and software components, including an A.I. system for the acquisition and automatic recognition of underwater images in different application contexts (e.g. different lighting conditions, different monitored organisms). The management of the logical and temporal flow of the components is optimized to reduce the energy consumption of the device. Some prototypes have been used in different biological monitoring contexts, within different scientific projects.

Possible Applications

  • Monitoring of the fauna biodiversity;
  • Monitoring of the gelatinous zooplankton (e.g. early-warning for jellyfish invasion);
  • Monitoring of the mucilage;
  • Monitoring of commercial fish species;
  • Monitoring of underwater geological structures (volcanoes, hydrothermal vents, landslides);
  • Wreck discovery and recognition.


  • Stand-alone and autonomous device;
  • Reduced size;
  • Reduced development costs (it allows extended areas monitoring);
  • Allows monitoring extended in time (> 12 months);
  • Can be installed either on fixed or mobile platforms;
  • Monitoring of remote and inaccessible areas.