Research from Stony Brook University suggests that tourist photos can aid in monitoring ecological changes in Antarctica. A team led by Heather Lynch, PhD, has developed a computerized method to analyze penguin photos taken by tourists. This technique helps determine the location of penguins over time and provides insights into changes in their abundance and distribution.
Lynch explains the potential of using tourist photos: “There are far more tourists in Antarctica than scientists, and virtually everybody has a camera in their pocket and many take photos of penguins.” The challenge was to extract information about where the penguins were located when the photo was taken, even without additional data.
The research utilized satellite imagery and 3D computerization to identify where a camera was positioned when taking a photo. This allowed researchers to estimate both the location and orientation of the camera. They then used an AI model called the Segment Anything Model to delineate penguin colony boundaries within photographs.
Identifying these boundaries is complex due to gradual edges with lower-density nesting near them. The process also involved georeferencing, which is challenging in Antarctica due to its lack of distinct features often used for image matching.
By creating a 3D model using satellite images draped over digital elevation models, researchers could pinpoint camera locations. They combined this with boundary data from the Segment Anything Model to locate penguin colonies precisely.
Lynch notes that this computational technique can be compared with other processed images to track changes over time: “In theory, this information gathered by the computational technique can be compared to other similarly processed images of the Antarctic to see how penguin colonies are changing over time.”
While satellite and aerial images are commonly used for tracking landscape changes, they are not always available. Tourist photos could expand data for long-term environmental monitoring significantly.
The research showed promising results but faced challenges due to variations in image quality and dynamic landscapes. Researchers believe it offers “a straightforward and effective tool for the georegistration of ad-hoc photos in natural landscapes.”
This interdisciplinary effort involved collaboration between ecologists, computer scientists, mathematicians, and geologists at Stony Brook University.
Alerts Sign-up