The surface infrared spectra of coated paper with different levels of calcium carbonate and SB-latex in the
coating layer were measured using ATR-FTIR. The relations between different levels of CaCO3 and SBlatex
and their respective characteristic peak areas were then established. The results show that the
general regression neural network (GRNN) model can be used to estimate the CaCO3 and SB-latex
contents in coatings of coated papers. The maximum errors for calcium carbonate and SB-latex were only
3.32% and 3.39%, respectively, and occurred under conditions that are at the extreme of what might be
encountered in an actual production facility.