Typical color image processing techniques use profiles consisting of sparse multi-dimensional lookup tables that interpolate between adjacent nodes to prepare an image for rendering. It has been suggested that spectral image processing could make analogous use of these techniques. A lookup table for spectral matching will relate spectral correlates to values related to printer digital counts. Previous research has shown that within a multi-colorant printer’s digital gamut there is often a large amount of spectral redundancy. Thus, without special consideration, it could be possible for adjacent nodes in a lookup table to have spectral consistency but digital count inconsistency. The colorimetric impact of lookup table interpolation within such situations was investigated.
Edward F. Hattenberger, Mitchell R. Rosen, Noboru