Color transformation modeling for printed images using interpolation based on barycentric coordinates

Publication Year:
Usage 7
Downloads 4
Abstract Views 3
Repository URL:
Gennetten, K. Douglas
Color printing
thesis / dissertation description
This document is a report on the research of a potentially superior method of color image transformation for producing rapid and accurate printed color output from a digitized color image. The primary objective is to reduce the complexities, inaccuracies, objectionable artifacts, and processing time common to current methods of color modeling. A new method of vector-corrected mathematical modeling combined with improved color space interpolation is studied. This achievement allows for rapid automatic closed-loop color-match calibration between image capture and output devices. Currently, digital image creation or manipulation systems rely on image proof printing which is slow and often of poor color fidelity or proprietary. The ideal system would provide optimum color reproduction fidelity and rapid proof printing at a low cost. Existing systems often rely on mathematical models for image con version for printing. These models are usually a bottleneck in the proofing process and are not accurate enough for many applica tions. Increasing their accuracy rapidly increases processing time to the point of impracticality well before graphic arts quality levels are achieved. A common solution to this problem in larger systems is a massive and tediously generated color look-up table based on actual measured print samples. This method is costly and does not readily accommodate printing process changes such as paper grade or ink color. Attempts to reduce the size of these look-up tables and the large quantity of required sample measurements have been disappointing. In this thesis, new methods will be reported which should allow practical small-system color proof printing with excellent color fidelity and rapid processing. By eliminating common problems associated with color space interpolation, these new methods make closed-loop control practical with a relatively small quantity of sample measurements which can be automatically printed, scanned, and incorporated into conversion processes.