Printer inversion algorithms and printer inversions with reduced measurements
“I, Kishore Narreddy, hereby deny permission to any individual or organization to reproduce this thesis in whole or in part.”
Physical copy available from RIT's Wallace Library at Z258 .N37 2003
The purpose is to find out an efficient algorithm to compute an accurate printer inverse map with structured input nodes. Several algorithms that are used to compute the inverse of a forward printer map are investigated. The forward printer map models the printer by mapping points in the printer's input color space to points in the printer's output color space. The inverse of this forward map is required to convert input color specifications in a device-independent color space to a color in the printer's device-dependent color space before being presented to the print engine. The accuracy of the inverse printer map directly affects the accuracy of the reproduced colors. Therefore, any measured change in the forward printer map requires re-computation of the inverse map if accurate and consistent color reproduction is to be maintained. An efficient and accurate method of computing the inverse map could be used in an automatic color correction system.
Four algorithms for computing the inverse of the forward printer map are studied in this thesis project. These are the Shepard's, Moving Matrix, Conjugate Gradient method, and Iteratively Clustered Interpolation (ICI) algorithms. The algorithms are implemented in C, Matlab and simulated in order to benchmark their relative accuracy, speed, and complexity. The simulations show the ICI algorithm to be the fastest, most accurate and less complex at computing the inverse map. A new Gamut mapping algorithm was derived and implemented in the lines of preserving the lightness and hue of the out of gamut nodes.