Skip to content

Attribution

We are grateful to all contributors who have helped improve Habitat-Mapper with code and data contributions, bug reports, testing, and feedback. Thank you!

Development and Support

Habitat-Mapper was developed by the Hakai Institute, a marine research organization based in British Columbia, Canada. The Hakai Institute led the development of this software package and contributed the majority of machine learning training data used to build the kelp detection and mussel and gooseneck barnacle detection models.

Have questions or feedback?

Send us an email or file a GitHub issue.

Training Data Contributors

The Nature Conservancy of California

The Nature Conservancy of California generously provided drone-based kelp imagery used to help train the kelp-rgb and kelp-rgbi models. This high-resolution aerial data was instrumental in improving model accuracy and robustness.

Dr. Katherine Cavanaugh

Dr. Katherine Cavanaugh et al. contributed valuable model development experience as well as expert annotations of Planet Labs satellite imagery that were used to help train the kelp-ps8b model1. We are incredibly grateful for her contributions and support of this project.

Model Contributors

Dr. Mohsen Ghanbari

Dr. Mohsen Ghanbari et al. developed the kelp-skema models as part of a post-doctoral project at Spectral Lab. These models were integrated into Habitat-Mapper from the original code, available on GitHub at https://github.com/m5ghanba/skema/.

Citation

If you use Habitat-Mapper in your research, please cite the software:

Denouden, T., & Reshitnyk, L. Habitat-Mapper [Computer software]. https://doi.org/10.5281/zenodo.17203205

@software{Denouden_Habitat-Mapper,
author = {Denouden, Taylor and Reshitnyk, Luba},
doi = {10.5281/zenodo.17203205},
title = {{Habitat-Mapper}},
url = {https://github.com/HakaiInstitute/habitat-mapper}
}

  1. Cavanaugh, K.C., Cavanaugh, K.C., Berberian, L.A. et al. High-resolution planet Dove data identify local drivers of kelp canopy persistence. Commun Earth Environ (2026). https://doi.org/10.1038/s43247-025-03134-y ↩