Automatic Picture-Matching of Crested Newts

Authors

Magnette G., Didry Y., Mestdagh X.

Reference

Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), vol. 12983 LNCS, pp. 329-334, 2021

Description

This article highlights a group of image processing, deep learning, and pattern matching techniques that can be used together in order to automatically identify different specimens of newts from a single species (Triturus cristatus Laurenti 1768). First, each image of newt will be: augmented, segmented, straightened. Then, patterns of images will be detected and compared between each other, allowing the differentiation of newts living in selected areas.

Link

doi:10.1007/978-3-030-88207-5_32