Artificial Intelligence: Detecting Marine Animals with Satellites
Working closely with the Alaska Fisheries Science Center, the Naval Research Laboratory and the Bureau of Ocean Energy Management, we are developing machine learning algorithms to detect marine animals from very high resolution satellite imagery.
Monitoring whales is of broad interest to governments, academics, and industry around the world. Scientists employ a variety of research platforms (aerial, vessel, passive acoustic) to monitor the abundance and distribution of marine animals and update stock assessments. Different survey platforms have their own strengths and challenges. Adding the detection of whales from satellites to augment our current platforms will allow additional flexibility as we seek to meet our mission requirements in the most effective and efficient manner possible. Several recent publications have demonstrated the technological feasibility of identifying whales from very high resolution (VHR) satellite imagery. NOAA is exploring the feasibility of an operational system using an AI pipeline in the cloud in a broad collaboration that includes the Alaska Fisheries Science Center, the Naval Research Laboratory, the Bureau of Ocean Energy Management, and the British Antarctic Survey.
Progress to Date
In 2020, we formed our network of collaborators and took a deep dive into the nuances of handling VHR satellite imagery. Very high resolution satellite imagery from WorldView-3, WorldView-2, and GeoEye satellites has been tasked over seasonal aggregations of North Atlantic right whales and the Cook Inlet beluga whale. The Naval Research Laboratory has begun to automate the workflow with a variety of Python scripts to query Maxar’s imagery database, address sun-glint phenomena and band offset, tile, and to detect objects of interest via spectral statistics.