Fluorescence microscopy is an established technique for qualitative an
d quantitative surface imaging. Because many ''real world'' surfaces o
f interest are optically rough (vertical variations on the order of 10
0 mu m or more) it is of great interest to minimize the amount of spec
tral information lost because of depth-of-focus variations. Multivaria
te statistics on normalized spectra provide an ideal method for extrac
ting information from such data. The following study examines nylon 66
gel on production pack screens via multivariate fluorescence imaging.