Recognition of Individual Zebrafish Using Speed-Up Robust Feature Matching

Animals' tagging has been widely used to identify their individuality using physical methods. In small swimming animals (e.g. zebrafish), however, physical tagging is considered a painful, costly and impractical. This paper proposes a new tagging method for zebrafish that is based on Speed-Up Robust Feature (SURF) matching. In this method, a set of local features is extracted from a sequence of image frames collected through a computer vision system. The extracted set of features for each free-swimming fish is then compared with pre-extracted sets of features, stored in a database, using the SURF matching method. Feature vectors through SURF are formed by means of local patterns around key points, which are detected using a scaled-up filter. The performance of the proposed tagging method is assessed experimentally using six free-swimming zebrafish. The obtained results demonstrated an average accuracy of 90% which obtained with a matching-features threshold of 15%. These findings are promising towards developing a painless, cost-effective and practical animal tagging system for zebrafish.