Hi Geospatial Enthusiast!

Would you like to predict the whole region AGB (Above Ground Biomass) with your field sample data? I will show you how on Earth Engine and Tensorflow (Keras) in this video.

GEE Script: https://code.earthengine.google.com/e...
Colab script: https://colab.research.google.com/dri...

I will use Landsat composite imagery on a West Kalimantan, Indonesia region. Then I will produce multiple spectral indices as additional variables. I am also adding the elevation data from SRTM as variables. All of the variables are scaled to 0 1 because it is much easier for deep learning to learn using that scale.

I am using a simple dense layer to help with the regression. Also in the Earth Engine, I am using the random forest regression to predict AGB too.

I hope this video helps with your project and research.

Email: [email protected]
Linkedin:   / ramiqcom  
GitHub: https://github.com/ramiqcom