In the early 1990s, South Africa became the first country worldwide to introduce great white shark conservation but lack of a standardised procedure for collecting data to help estimate population numbers has hampered efforts to manage and protect these animals. Until now, that is.
A marine biologist, a mathematician and a software developer from South Africa’s Stellenbosch University have joined forces to create an image-recognition software package that maps dorsal fins in order to identify individual sharks.
Similar to fingerprint identification, the Identifin software uses an algorithm technique (called dynamic time-warping) to match the trailing edge of a dorsal fin. It compares a semi-automatically drawn trace of the back edge of the fin with images in the database, which are then re-arranged and ranked by probability of a match.
