Computer vision and machine learning
Emodnet Bathymetry is a project founded by the European Union’s under DG-MARE and EASME and pretends to understand the topography of the European seas.
The project is coordinated by SHOM, it started on January 2017 and will finish on December 2018, recently the consortium obtained a 2 year project renewal where Coronis is also invited to continue working on it.
The purpose of Coronis in EMODnet Bathymetry was to help in updating the visualization of the global DTM to 3D. In order to achieve this goal, we wanted to use web-based visualization applications. In these applications, huge amount of data needs to be passed through the net and the rendering happens on the user side. Thus, it is important to keep a balance between the amount of data to transfer and the amount of effort required for rendering it.
Consequently, Level of Detail (LOD) techniques, able to change the complexity of the displayed data based on the point of view required by the end user, are desirable in our context. These techniques focus on rendering the part of the world falling in the user’s frustrum with a complexity that adapts to the distance from the viewer or the projected screen size (i.e, the complexity of the data should adapt to the perception of the user given a point of view).