Sea Slugs species recognition using images
Master of Applied IT
Client company:Blauwtipje
Iustin-Mihai Tescan
Joris Kamminga
Project description
Blauwtipje aims to create a way for people to identify sea slugs they may have
encountered underwater. They want to simplify sea slug identification by
utilizing technologies which can help the process.
Context
Blauwtipje is a social citizen platform based in the Netherlands where sport divers, beachcombers, and all other observers of nudibranchs (or casually called Sea Slugs) can enter their observations here and thus keep a personal sea slug logbook. Because of this Blauwtipje has collected quite a large number of pictures and observations of various species of sea slugs in different locations. The database is used anonymously for scientific research on the dynamics of nudibranch biodiversity in the Netherlands.
Results
The outcome of this project is the creation of a Convolutional Neural Network (CNN) model capable of accurately recognizing 65 different species of Sea Slugs. An infrastructure for the automation of continuous training and deployment of the model was also developed.