Tourists flock the Falkland Islands for their stunning landscape, beaches, and magnificent bird life. But the scientists who routinely need to visit the rocky, uninhabited outer islands to keep an eagle eye on bird colonies often do that with an albatross around their neck (well, not literally).
Counting bird populations manually is a laborious process. Often, the presence of humans can disrupt the birds’ breeding and nursing behaviors. And since bird colonies are typically large and densely interspersed, population counts need to be repeated as well.
The Falkland Islands are home to world’s largest colonies of black-browed albatrosses and second-largest colonies of southern rockhopper penguins. Keeping tabs on their populations is important for wildlife conservation.
And so, scientists are turning to drones and artificial intelligence (AI) as an alternative to traditional, on-the-ground methods. And a new study conducted by Duke University and the Wildlife Conservation Society (WCS) finds that these methods can be just as effective, while reducing costs, labor, and the risk of human error.
Madeline Hayes, a remote sensing analyst at the Duke University, who led the Falkland Islands study, says:
Using drone surveys and deep learning gives us an alternative that is remarkably accurate, less disruptive, and significantly easier. One person, or a small team, can do it, and the equipment you need to do it isn’t all that costly or complicated.
A bird’s-eye view to count birds
For their study, the scientists analyzed more than 10,000 drone images of mixed colonies of seabirds in the Falkland Islands.
Using a combination of drone data and AI algorithms allowed them to correctly identify and count the albatrosses with 97% accuracy and the penguins with 87%. Overall, the automated counts were within 5% of human counts about 90% of the time.
The analysis was done using a convolutional neural network (CNN), a type of AI that employs a deep-learning algorithm to analyze an image and differentiate and count the objects it “sees” in it. David Johnston, director of the Duke Marine Robotics and Remote Sensing Lab, explains that the CNN is loosely modeled on the human neural network, meaning that it learns from experience. He tells:
Essentially, the computer identifies different visual patterns, like those made by black-browed albatrosses or southern rockhopper penguins in sample images, and over time it learns how to identify the objects forming those patterns in other images such as our composite photo.
The scientists are hopeful that this new approach of using drones and AI to count seabirds will have a significant impact on marine conservation. It would increase the ability of ecologists to monitor the size and health of seabird colonies worldwide, as well as the health of the marine ecosystems they inhabit.
Photo: Madeline Hayes, Duke University
FTC: DroneDJ is reader supported, we may earn income on affiliate links