The world’s leading publication for data science, AI, and ML professionals.

Finding satellite images for your Data Science project

Sentinel Hub, one of the best resources for satellite images for your Data Science project

Photo by Tobias on Unsplash
Photo by Tobias on Unsplash

There are a lot of things you can do with satellite images nowadays, be it to use in your geographical analysis system or to aid your statistical analysis on R/Python or any other programming language.

The first question you will (and should) ask yourself though is: where can I find these satellite images? The answer(s) to this question is right here. I’ll show you the best resource where you can find high-quality satellite images from anywhere in the world.

Next, I’ll show how to obtain these images from a certain region. I’ll work with Oporto – a beautiful coastal city in Portugal 🇵🇹 , as my region of choice because it’s where I live. The best source to obtain satellite imagery IMO is:

Sentinel-hub EO-Browser3

  1. Sign up for an account, it’s easy and quick. Sign in with that recently created account and you’ll see this screen:
  1. Zoom out and move the map to the place you wish to obtain the satellite images. Click "Search" and you’ll see all images that cover the area you selected.
  1. Choose the image that serves you best, for instance, the one with the least cloud cover is usually the best to perform any kind of analysis. When you found your favorite, click on this button to access download links:

Click on the second link to start the download. It’ll open a new page and ask for a user and password. It’s the same information you used to login to the website a few minutes ago. Enter the user and password and the download will start right away.

And that’s it. In a few minutes, you’ll have the satellite images you need and you can open them in your software of choice, be it SNAP, ArcGIS, RStudio, or any other you want.

Some other functions and notes on Sentinel Hub

  • You also have the option to choose images from another satellite. Landsat is another very popular choice to obtain your images (and it’s available here as well)
  • A date filter will allow you to choose a month or a period in particular:

You can work with maps directly on Sentinel Hub’s interface, to do things such as:

  • Place markers, measure distances:
  • Calculate areas:

And also create timelapse videos or animations with the maps you’ve chosen.

Final thoughts

That’s it for the basic stuff you need to know about Sentinel Hub. I believe it’s one of if not THE best resource for downloading satellite images for your Data Science projects.

Check a few good articles by Abdishakur, Adrian Serapio, and Anubhav Pattnaik on Geospatial Analysis to inspire yourself for your next project:

External Geospatial Data: Types and Use-cases

Simple, Visual, Geospatial Analysis Towards Pandemic Readiness

Land use/Land cover classification with Deep Learning

Check out Sentinel Hub whenever you have a chance, find the images you need, and let me know what you think in the comment section!


Related Articles