We know the seas are changing. Temperatures are rising and  many marine environments across the planet suffer from acidification and plastic- and chemical pollutions.

UN Sustainable Developement Goal number 14 states, that we should preserve and protect the earth’s marine and coastal ecosystems, but in order to do so, we must know more about what really goes on deep in the sea, which is hard to monitor.  Manual data-collection at sea is expensive and essentially impossible to do for longer periods of time, but the research project “Marine analytics using computer vision” aims to change this.

On the pellets of the Limfjord-bridge - nine meters under the surface of the unruly waters of the inlet - three underwater cameras are installed. They now send a live video-feed that tracks every movement and every change in the condition of the water, and will be used to develop algorithms that can identify species and activities in waters with changing visibility.

The project researchers collaborate with Danish and American companies Ambolt and Kitware that specialize in applying AI/computer vision in different domains. The aim of the project is to develop a system, that automatically analyzes the underwater conditions.

That way we can produce valuable datasets that are both vaster and much more varied than the data, you are able to collect otherwise, and this can prove useful for marine biologists and others who work to understand and solve the problems of the seas.

Project funded by The Independent Research Fund in Denmark.

• Læs også artiklen Kunstig intelligens skal indsamle mere og bedre maritimt data til at forstå miljøproblemer af Mette Lynge Hansen, Danmarks Frie Forskningsfond

Contact

Professor Thomas B. Moeslund and Ph.D. Malte Pedersen, Department of Architecture, Design & Media Technology.