Newbie question: data aggregation & filtering with SNAP Desktop


I’m working on a class project for a data mining / analysis class at Cleveland State University in the United States. We’ve decided to use the the SNAP Desktop toolbox for our analysis, so I was wondering if I can ask for some help?

What we’re trying to accomplish: the goal of the project is to analyze the vegetation levels (like the amount of tree cover) in various regions (municipalities). So ultimately, I think we’d need run an aggregation function on the pixels for each region to produce a score for that region, perhaps using various bands (so multiple scores).

What we accomplished so far:

  • We found and downloaded the S2 data for the regions we’re looking for. There are two separate downloads that cover our contiguous area to be analyzed.
  • We were able to open the S2 data in the SNAP toolbox and can zoom in and toggle the various bands. We also figured out how to generate RGB satellite imagery for the data we downloaded.
  • The area is broken out into multiple regions (there’s something like 40 regions in the area). We got our regions as a shapefile and were able to import them into the SNAP toolbox successfully. (Vector -> Import -> ESRI Shapefile)
  • We can display the regions as nicely overlaying the satellite imagery.

What we need help with:

  • We have no idea how to aggregate pixel / color data per region (vector boundary).
  • We don’t know how to combine multiple satellite “products” into one contiguous image for analysis.
  • Any other tips / tricks you can think of that can help us? We are brand new to this field and are finding this type of analysis to be challenging!

We would sincerely appreciate any help you can offer on any/all the points mentioned!

Thank you very much in advance for your help,
Greg Sherman