Skip to content

A deep-learning framework to map the benthic habitat distribution of the Mediterranean Sea.

License

Notifications You must be signed in to change notification settings

agimenezromero/CAMELE

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

10 Commits
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

CAMELE

Despite the mapping of several marine habitats has been successfully achieved through the use of side-scan sonar units, the cost and time-intensive nature of these methods present challenges in deploying continuous monitoring systems. We aim to address these challenges by combining the vast amount of remote sensing data available with recent advances in deep learning to develop a framework that can automatically map marine habitats from satellite imagery. CAMELE (Consensus for Automatic Marine Ecosystem Labelling and Evaluation) is a deep learning-based system to automatically monitor marine ecosystems from satellite imagery. This will enable the continuous monitoring of marine ecosystems at a fraction of the cost and time required by traditional methods and the opportunity to monitor larger areas.

Here you can find an example code to use CAMELE model to predict the underlying habitat distribution of two satellite images from the coast of the Balearic Islands. The models and the data can be found at zenodo.

Requirements

  • Python 3.X

Instructions (Linux)

  • Download the models and the data from the Zenodo repository and place them in the corresponding folders.

  • Create a suitable conda environment using the requirements.yml file

    conda env create -n requirements.yml
  • Activate the environment

    conda activate camele
  • Now you are ready tu run the Example.ipynb notebook. Have fun!