Visual-based solutions for an effective use of Autonomous Underwater Vehicles in environmental monitoring applications

The strategies developed in the work of this thesis can be divided into two topics: underwater navigation and decision-making based on environmental information. The first part of the thesis deals with navigation over P. oceanica meadows and describes the two different visual-based navigation solutions. The second part of the thesis concerns the autonomous boundary inspection of P. oceanica meadows, and in particular the modules that form the proposed visual inspection framework. The validation of the proposed methodologies has been performed through realistic simulations and experimental data collected at sea using an underwater robot.