Deutsches forschungszentrum für künstliche intelligenz gmbh

Published purpose

The collaborators aim to enhance an AI-based computer vision program so it can be deployed to monitor Canadian waterways and identify problem areas for plastic pollution in need of remediation. The proposed project takes Canadian (NRC) expertise in computer vision for the identification of ice and pairs that with German (DFKI) expertise in computer vision expertise in the identification of plastic to meet the Oceans Challenge Program's need for pollution remediation. The project takes an existing German computer vision program that can identify plastic in ice free waterways and applys it to Canadian waters. Because a computer vision algorithm trained to identify plastic in rivers could be confused by floating ice, it is seen as a necessary improvement to the model to train it to distinct between ice and plastic litter. The outcomes of the project can add benefit of helping to understand plastic waste situation in Canadian rivers.