Semi-automated software for dorsal fin photographic identification of marine species: application to Carcharodon carcharias

摘要

When dealing with large marine species, individual photographic identification plays a very important part in capture-mark-recapture studies. The success of this method is determined by the consistent correct identification of individuals, but as the number of images in the database increases, the task becomes increasingly time-consuming, affecting the accuracy of the matches. Although a few software packages are available, universally applied methods remain problematic in long-term studies, suggesting the need for species-specific software. This study presents new image recognition software tested on white sharks' dorsal fins and is based on a dynamic time warping algorithm. The software is specifically designed to improve the matching success of individuals, to standardise the data collection and to increase the overall accuracy when managing a large capture-mark-recapture database.