Spotting the difference: towards fully-automated population monitoring of African penguins Spheniscus demersus

A novel computer vision system demonstrates high accuracy and efficiency in non-invasive population monitoring of African penguins at Robben Island, South Africa. The system identifies individuals using unique chest plumage markings with a false identification rate below 1 in 10,000 comparisons and a genuine acceptance rate of 96.7%. It successfully monitors over 13% of the penguin population passing a camera, with potential for higher detection under favorable conditions. This approach outperforms traditional flipper banding methods and offers a scalable solution for fully-automated, non-invasive wildlife monitoring, as validated by field testing.

Sherley, R., Burghardt, T., Barham, P., Campbell, N., & Cuthill, I. (2010). Spotting the difference: towards fully-automated population monitoring of African penguins Spheniscus demersus. Endangered Species Research, 11, 101–111. https://doi.org/10.3354/esr00267