Reproducing Color Images Using Custom Inks

Overview

In previous work (Reproducing Color Images as Duotones), we developed a technique for printing full-color images using just two inks instead of the traditional four. The key idea was to allow the computer to choose the optimal inks for a given image based on the colors contained in that image. In this research, we extend our previous work to printing color images with n custom-selected inks, where n is chosen by the user according to budget constraints and quality concerns.

The fundamental change when we go from two inks to more than two is the shape of the gamut of printable colors. In three-dimensional color space, a duotone gamut is a curved surface, while for more than two inks the gamut is a volume. Thus, we need a completely different algorithm to map the colors in an original image to the print gamut. We also need a more accurate mathematical model of the print gamut than we did for duotones; with more than two inks, users demand an accurate reproduction. In addition, finding the amounts of inks needed to reproduce a given color was a fairly simple mathematical problem for two inks, but a difficult one (with multiple solutions) for more than two inks. Finally, we need a robust combinatorial optimization algorithm for selecting the inks that best reproduce an image.

We explored these issues related to printing with multiple custom inks, and addressed them with new algorithms and physical models. The article and dissertation below describe our techniques and present some printed examples demonstrating their promise.

Personnel

Eric J. Stollnitz <stoll@amath.washington.edu>

Victor Ostromoukhov <victor@lspsun16.epfl.ch>

David H. Salesin <salesin@cs.washington.edu>

Publications

Eric J. Stollnitz, Victor Ostromoukhov, and David H. Salesin. Reproducing color images using custom inks. In Proceedings of SIGGRAPH 98, pages 267-274. ACM, New York, 1998.

[Acrobat file, 209 Kb]

[gzipped PostScript file, 511 Kb]

Eric J. Stollnitz. Reproducing Color Images with Custom Inks. Ph.D. dissertation, University of Washington, 1998.

[Acrobat file, 1.7 Mb]

[gzipped PostScript file, 5.3 Mb]