I'm trying to detect clouds and their shadows on Sentinel2 images with SentinelHub service.

For clouds detection, SentinelHub team has developed a good tool but I would like to detect cloud's shadow too.

For the moment I use B08 band (NIR) with a threshold of 0.15, it's an empirical method I created. But maybe it exists a better way... Somebody have an idea?

  • If you have historical data about wheather, and especially cloud height and you know the date/time/location the images are taken from, you could calculate based on the detected clouds where their shadow should be. So a 3D spatial approach. – LMB Nov 20 '18 at 11:40
  • Perhaps even underlay it with a DEM for more accuracy. – LMB Nov 20 '18 at 11:48

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Browse other questions tagged or ask your own question.