I am working on a project for San Jose, CA and I am trying to determine the change in the urban tree canopy and am using 0.6m NAIP images. I did the change detection and then realized I needed to classify the NAIP first to yield a more logical output. I tried out the DL package with no luck following the instructions. I also have a support ticket open to try and help as I'm on a time crunch but so far that isn't helping.
See attached change dection without classification. This output does not yield useful results. I need the classification to eliminate things like buildings and parking lots.
I don't know about your specific issue but, because you have a time crunch, perhaps you could use this NOAA dataset instead of trying to build one off of NAIP.