Abstract: The Sea Otter Nose Matching Program, or SONMaP, was developed to identify individual Alaskan sea otters using a blotch-pattern recognition algorithm based on the shape and location of lightly colored nose scars. The program ranks all the images in order of similarity, most similar first, with six images displayed at a time. The user then selects the final match. In this study, the performance of the SONMaP program was tested using images of otters that had been previously matched by visually comparing every otter in a catalog of 1,638 animals. After running the images through SONMaP, they were classified as BEST, AVERAGE, or WORST based on whether the correct match was within the first 10%, 11 to 50%, or 51 to 100% of images in the catalog, respectively. In 48.9% of the previously visually matched images, the program accurately selected the correct image in the first 10% of the catalog, which compares favorably with other computer-assisted photo-identification studies of marine mammals.
Key Words: SEA OTTER; ENHYDRA LUTRIS; PHOTO-IDENTIFICATION; COMPUTER-ASSISTED MATCHING; SONMAP; ALASKA
Document Type: Research article
Page Numbers: 349 - 358