Sea2space is a project focused on developing adaptive mapping algorithms for both underwater and planetary science robotics. Many phenomena scientists want to observe in the natural world are governed by various geological and biological processes at a range of scales. We aim to use computer vision techniques to identify different terrain and habitat types at the macro scale (e.g. coral reefs, rock types) to help target exploration of phenomena at the microscope (e.g. locating subsurface Martian water or localising diseased coral). In Svalbard we are gathering invaluable datasets on planetary-like environments to be used for the development of these algorithms. 

Project Contributors: Jessica Todd, associated groups: Human Systems Lab, AeroAstro, WHOI Autonomous Robotics and Perception Lab, Woods Hole Oceanographic Institute