SERVICES

What We Do

SPACEWHALE is a novel method for detecting whales in very high-resolution (VHR) satellite imagery which can monitor regions in the high seas where traditional boat or aerial-based surveys are difficult to execute. We merge the latest in space-based and Deep Learning assets with a team of expert biologists to offer clients a range of products including abundance and distribution estimates.

Abundance


We provide abundance estimates for large whales on a per-survey basis which can be used to generate population trends over time.

Distribution


Using advanced modelling techniques, we create distribution maps of target species in ESRI-compatible raster formats.

Bespoke analyses


Working with our clients we offer bespoke analyses to help guide conservation and advanced ecological studies.

Projects

Hawaii / Peninsula Valdes

Our pilot project, which was run in 2018 – 2019, tested the possibility of detecting whales from satellite imagery using a deep learning algorithm. Using a standard classification technique, we were able to identify 100% of the whales in the images.

Find more information at “Our Publications”

Humpback whales (Megaptera novaeangliae) photo: Michael Schauer
Fin whale (Balaenoptera physalus) mother and calf photo: Caroline Hoeschle
Projects

Pelagos Sanctuary

We applied the SPACEWHALE faster R-CNN object detection method to the Pelagos Sanctuary previously studied by the British Antarctic Survey. The SPACEWHALE algorithm succeeded in detecting an additional 13 whales in satellite imagery that were overlooked by a manual review process.
Projects

Bay of Biscay

During September 2020, we flew digital aerial surveys in the Bay of Biscay and captured satellite imagery simultaneously. We are investigating the overlap between traditional aerial surveys and cutting-edge semi-automated techniques using satellite images.
Humpback whale (Megaptera novaeangliae) photo: Caroline Hoeschle

Mapping Whales From Space - Our Publications

We are actively publishing the results from our work in peer-reviewed journals. These manuscripts remain some of the first publications that demonstrate deep-learning for detecting whales from satellite imagery.

Open our publications as PDF by clicking on the buttons below or find more information for Borowicz et al., 2019 (https://doi.org/10.1371/journal.pone.0212532) and for Höschle et al., 2021 (https://doi.org/10.3390/s21030963)

Aerial Trained AIPotential of Satellite Imagery
DB+
Our Team

Research Team

Caroline Höschle

Caroline Höschle

Project lead |Biologist

Dr. Amel Ben Mahjoub

Dr. Amel Ben Mahjoub

Image recognition specialist

Dr. Grant Humphries

Dr. Grant Humphries

System architect | Data scientist

Vladislav Kosarev

Vladislav Kosarev

Programmer

Dr. Georg Nehls

Dr. Georg Nehls

Managing director BioConsult SH

Marc Schnurawa

Marc Schnurawa

Spatial data analytics

Dirk Renner

Dirk Renner

Flight organisation and Data management

Kai Christensen

Kai Christensen

Review / ID Team coordinator

Martin Scott

Martin Scott

Commercial director HiDef

Dr. Katharine Keogan

Dr. Katharine Keogan

Marine data scientist

Our Team

Review Team

Theo Jensen

Theo Jensen

Review

Volker Schlemm

Volker Schlemm

Review

Evlin Mirza

Evlin Mirza

Review

Karo Galoyan

Karo Galoyan

Review

Morsal Aslami

Morsal Aslami

Review

Lennart Rienitz

Lennart Rienitz

Review

Farhad Ataee

Farhad Ataee

Review

Gunnar Wolff

Gunnar Wolff

Review

Carsten Klein

Carsten Klein

Review

Sardar Mirza

Sardar Mirza

Review

Nevard Saroyan

Nevard Saroyan

Review

Contact us for a quote

By submitting your request, you consent to the processing of your data for the purpose of processing your request (data protection declaration and revocation instructions).

Office

Schobueller Str. 36,
25813 Husum, Germany