Extract Clouds from NAIP Imagery?
Joan
We could brainstorm a lot of creative ideas, but in the end my advice would be "don't try".
First, NAIP has very few clouds. It took me quite a while to find this pair:
Second, NAIP is so heavily processed, it will be very hard to come up with a tool (even using state of the art machine learning or "deep learning" methods) that can recognize the shapes and pixel values of clouds. The cloud above was on a seamline between 2 images, and got clipped such that it has a rectangular edge! You can see them in this map http://arcg.is/1KnO1K and note that the two white buildings in this scene have brighter pixel values than the clouds. If you try to highlight all the "white" pixels, I'm guessing you'll have 1000:1 false positives vs. actual clouds. (Note you have to sign in to ArcGIS Online to see the NAIP image service in the map link)
With NAIP, you also don't have sensor location and image time metadata to attempt to correlate candidate clouds with corresponding shadows. So, applying this sort of logic would be complicated but *possible* with satellite data (although not certain it would succeed), but you can't even attempt it in NAIP.
On top of the challenge of reliably finding clouds and eliminating false positives, you'd have another significant challenge in creating a polygon. In the above example, where would you draw your ideal polygon boundaries?
My key question is "why?". If you're performing some form of auto classification and concerned about errors, I think clouds are the least of your worries. Again, there are very few in the NAIP imagery (at least everywhere I've looked), so statistically any cloud errors will be negligible compared to other error sources. The geometric quality is great, but the radiometric quality of NAIP is not suitable for doing any but the simplest automated analysis - gross delineation of vegetation using NDVI can be somewhat successful, but clouds would be a very minor issue.
NAIP is great for manual interpretation (and you can't beat the price), but if you're attempting to automate some analysis, I'd be very cautious about expectations. If you can share information here about your higher objectives, we might have some useful advice.
Cody B
