Hi-- I have a thought (still new to this), and wanted it to run it by the experts--please let me know if I'm going about this correctly! I'm doing some archaeology in the desert, and trying to study erosional processes, etc.  I have some quickbird imagery--2.5m multispectral and .6m panchromatic.  Individual desert shrubs can be picked out on the panchromatic imagery, just generally inferred on the ms.  What I want to do is create a raster showing vegetation density, but sometimes the vegetation and shade look similar.  My thinking is this: Create a Gram-Schmidt pan sharpened image with all 4 MS bands.  Minimize the shade through the use of a band ratio, then run a classification, possibly using the rule images to determine a threshold.  I'm not sure how to use endmembers, fraction images, etc. to isolate the shade, though.  Does this sound like I'm going about it the right way? Thanks! Steve