Massimal

Mapping of Algae and Seagrass using Spectral Imaging and MAchine Learning

Publications and presentations

Datasets

Skjelvareid, M., Rinde, E., Hancke, K., Blix, K., Hoarau, G. G., Kile, M. R. (2025).MASSIMAL: A multimodal dataset for coastal habitat mapping in Norway [Data set]. Norstore. https://doi.org/10.11582/2025.00041

Hyperspectral, multispectral and RGB datasets from the project have also been published via the SeaBee Geo-Visualization Portal.

GitHub repositories

  • massimal-dataset: Documentation and example usage of Massimal dataset (see above) 
  • massipipe: Pipeline for calibration, processing and georeferencing of hyperspectral images. 
  • vidtransgeotag: Tool for extracting geotagged images from video and position data streams.
  • massimal: Collection of code and notebooks used for processing Massimal data.

Conference papers

IGARSS 2023 - Mapping marine macroalgae along the Norwegian coast using hyperspectral UAV imaging and convolutional nets for semantic segmentation

Conference presentations

NORA (Norwegian Artificial Intelligence Research Consortium) conference 2023 - Presentation of Massimal project and preliminary results

Posters

Northern Lights Deep Learning 2022 - Poster presentation

International Temperate Reefs Symposium 2023 - Poster presentation

Marine Geological and Biological Habitat Mapping (GEOHAB) Conference 2023 - Poster presentation

Marine Geological and Biological Habitat Mapping (GEOHAB) Conference 2024 - Poster presentation: Mapping shallow-water maerl beds with UAV hyperspectral imaging

Master's theses

Silje B.S. Grue (2022) - Machine Learning for Classifying Marine Vegetation from Hyperspectral Drone Data in the Norwegian coast