15  Featurization demo

This chapter is in early draft form and may be incomplete.

15.1 Overview

This demonstration will show you a few key concepts

15.2 Demonstration code

Below is a link to a Jupyter notebook intended to demonstrate practical featurization of satellite data for use in MOSAIKS. The notebook will guide you through the process of preparing imagery, including:

  1. Building a pytorch model from scratch

  2. Using an out of the box solution with torchgeo

  3. Building a torch dataset and dataloader

  4. Featurizing satellite imagery

  5. Saving features to disk

  6. Visualizing features

Click the badge to run the demonstration!

↓↓↓↓↓↓↓↓↓↓↓↓↓↓↓↓
Open In Colab
Remember to click File -> Save a copy in Drive to save any changes you make.

Or to view a static version of the code on GitHub, click the badge below.

For instructions and tips on using Google Colab, please refer to Chapter 1.

15.3 What’s next?

Now that we have covered the basics of featurizing satellite data, we will move on to the next section, where we will discuss modeling.

Looking forward

This is the end of the feature section. Next, we will move on to the modeling labels with satellite features.