The Napoleon, or humphead, wrasse (Cheilinus undulatus), is a giant and ornately decorated reef fish which can grow over 6 feet long and live for over 30 years. Unfortunately, these rare fish are a coveted delicacy in live fish markets throughout Asia, and the high volume at which they are traded has landed them a spot on the CITES endangered species list.
To keep them from going extinct, stricter regulations must be implemented and enforced surrounding the collection and trade of Napoleon wrasses. Unfortunately, fish trading networks are complicated, and it can be difficult to track where fish come from and if they are being sold legally.
While all fish imported to major markets may be inspected, those within the live fish trade cannot easily be tagged or marked, making it difficult to track these fish from the point of capture through to their final sale point. But a pair of researchers from the University of Hong Kong decided to try a new technique in lieu of a physical fish barcode: facial recognition software. Napoleon wrasses have complex facial patterns that the researchers hoped could be used as a unique identifier for individuals.
First, they had to test if the facial markings were unique and variable enough between fish to reliably identify individuals, and that they did not change over time. Then to ensure that the identification system was feasible, researchers developed and tested an app used to compare pictures of Napoleon wrasses at fish markets to those in a database of previously identified fish. And it worked – this study served as a proof-of-concept that facial recognition software could reliably identify endangered Napoleon wrasses sold in live fish markets.
Although not currently used on a large scale, there is real potential that the software and app developed could improve the monitoring and enforcement of legal endangered species trading. This approach could also apply to other animals that have distinct, readily identifiable markings.