Analyzing the Amazon Deforestation with Machine Learning and the Google Earth Engine – Part 1

Go back to programme

Analyzing the Amazon Deforestation with Machine Learning and the Google Earth Engine – Part 1

In the first part of the GeoAI Discovery Series, learn how to apply machine learning (ML) algorithms to analyse deforestation in the Brazilian Amazon. Using open geospatial data, in particular multispectral images from the Google Earth Engine data catalogue from ESA Sentinel and NASA Landsat missions, AI can show the loss of forests over time.

This AI for Good Workshop provides an introduction to the Google Earth Engine data catalogue and processing capabilities, and shows how to browse and access their petabyte data catalogue. We also introduce basic JavaScript for users to be able to manipulate and handle data processing at a certain level.

Going further into the Workshop, learn how to compute indices such as Normalized Difference Vegetation Index (NDVI) and how to analyze them in a time-series manner. All the showcased steps will lead to a preprocessed set of scenes which will be further used to apply a machine learning algorithm to quantify forest loss. Finally, the results and products will be evaluated against an external dataset.

For more information on the applied techniques, view the paper here.

Share this session
Scroll Up