A note on an automated system for matching the callosity patterns on aerial photographs of southern right whales

Main Article Content

Lex Hiby
Phil Lovell

Abstract

Photographs showing the callosity patterns of southern right whales (Eubalaena australis) are currently compared by eye to identify
individuals and monitor their occurrence within certain areas. This paper describes software designed to reduce the number of by eye
comparisons required to maintain each of the existing local photo-identification catalogues. The software is used to extract, from each
photograph, a viewpoint-independent description of the shape and location of each callosity which generates a parallel catalogue of extracts.
This is then compared with the description extracted from each new photograph to generate a list of similarity scores and thus highlight
likely matches. The software can also be used to compare the different catalogues of extracts with each other. Using a test set of 67
photographs of 23 whales taken from 1974 to 1986, the software reduced the number of by eye comparisons required to identify all
individuals by 93% when compared with a purely random search.

Article Details

How to Cite
1.
Hiby L, Lovell P. A note on an automated system for matching the callosity patterns on aerial photographs of southern right whales. JCRM [Internet]. 2020Oct.30 [cited 2021Jun.24];:291-5. Available from: https://journal.iwc.int/index.php/jcrm/article/view/278
Section
Articles