Global Mapper v25.0

Issues auto-ground classifying a cloud derived from an orthophoto

I have a pointcloud derived from an orthophoto and am trying to use the auto-ground classify function, but it keeps thinning out the cloud immensely.  I've ran the function with multiple specified colors that correlate to the ground and had the restrictions set all the way to the right on max 'fuzzy', but it still only picks up a minimal amount of the ground.  I understand the fact that every point only has one return, due to being ortho-derived, but I can't seem to figure out how to get around this.  All suggestions are appreciated.