PhD student
Funding: Ifremer
E-mail : louis-jouclas (at) ifremer.fr
E-mail : louis-jouclas (at) etud.univ-tln.fr
Bio
Graduate of ESIEA, specializing in embedded systems.
Offices
- M building, Université de Toulon, Campus La Garde.
- Centre Méditerranée – Zone Portuaire de Brégaillon CS20 330 – La Seyne-sur-Mer.
Supervision
- Vincent HUGEL, professor at UTLN (co-director),
- Vincent CREUZE, professor at University of Montpellier (co-director).
- Claire DUNE, associate prof. at UTLN (co-supervisor),
- Maxime FERRERA, senior researcher at IFREMER, team PRAO, IFREMER, la Seyne-sur-Mer (co-supervisor),
Topic
Vision-sonar-based exploration for autonomous underwater robots
The context of this topic is the exploration of an unknown natural or artificial 3D structure by an underwater robot. The objective is to design an autonomous navigation method for mapping the 3D structure to ensure complete photogrammetric coverage, while guaranteeing vehicle safety. Existing underwater SLAM (Simultaneous Localization and Mapping) methods have demonstrated that it is possible to locate and map the underwater environment using vision-only systems, or systems combining vision and sonar. However, in the first case, the robot was piloted by a human during exploration, while in the second, it was moved directly by a diver in a cave. We propose to combine progressive construction of the environmental map with vehicle guidance, so as to guarantee both complete coverage of the area explored and optimum shooting conditions for the photogrammetry system (contrast, sharpness, luminosity, colorimetry, etc.). The aim is therefore to generate safe trajectories, while adding the constraint of optimizing the photogrammetric result.