Identifying tiny planktonic critters in the Arctic Ocean using a submersible camera system and artificial intelligence

Members of Québec-Océan recently published a paper based on a submersible camera system called LOKI. Deployed in the high Arctic, LOKI took images of zooplankton.

BarretteJ_201308150663_WM copy
Deploying LOKI from the CCGS Amundsen ( Photo credit:Jessy Barette)

Using these high definition images, an automatic identification model based on machine learning was developed that can reliably turn these images into taxonomic information. The camera system’s high vertical resolution (<1m) in combination with a very high resolution taxonomic model, identifying developmental stages of copepod species, enables scientists to study the link between zooplankton species and their abiotic as well as biotic environment with much detail. Especially the coupling between zooplankton and their phytoplankton prey is of high relevance in the face of climate change in the Arctic.

The model was recently used together with in-situ lipid analysis from images, yielding some very interesting results, so stay tuned!

Sample images of LOKI (taken from Schmid et al., 2016)

To read the paper :

Schmid, MS, C Aubry, J Grigor, L Fortier (2016). The LOKI underwater imaging system and an automatic identification model for the detection of zooplankton taxa in the Arctic Ocean. Methods in Oceanography 15–16: 129–160.

Follow the project at

Arcticnet2013 leg1b-Cyril Aubry 90
Image of Peterman Glacier photographed from the CCGS Amundsen. LOKI Images from here were also used to develop the automatic identification model presented in the paper (Photo credit: Cyril Aubry).

Laisser un commentaire

Entrez vos coordonnées ci-dessous ou cliquez sur une icône pour vous connecter:


Vous commentez à l'aide de votre compte Déconnexion /  Changer )

Photo Google+

Vous commentez à l'aide de votre compte Google+. Déconnexion /  Changer )

Image Twitter

Vous commentez à l'aide de votre compte Twitter. Déconnexion /  Changer )

Photo Facebook

Vous commentez à l'aide de votre compte Facebook. Déconnexion /  Changer )


Connexion à %s