Using color management to automate the color reproduction of 3-D images procured via a digital camera/3-D scanner

The use of digital photography is migrating from the major applications in pho tojournalism to professional studio photography. Traditional service bureaus such as professional photo labs and prepress trade shops are adding digital imaging services to their film-based services. Also, businesses such as advertising agencies and publishers, who traditionally outsource work to service bureaus, are bringing digital imaging services in-house. State of the art imaging technology empowers users with new tools, but does not guarantee that the task of generating accept able image reproductions will be easier. The basic problem in the desktop color prepress environment is that each com ponent in this open system handles color differently. Miscommunication between devices results in user frustration with an unpredictable, inconsistent, and inaccurate color system. The solution to this problem is to assess one\u27s workflow and adopt a color management system (CMS). The purpose of CMSs is to help users maintain color integrity throughout their desktop system and to automate the color separation process. This thesis project investigated the possibility of applying a comprehensive CMS to automate the color reproduction of 3-D images procured with a digital camera. Automatic exposure by Leaf System\u27s Lumina digital camera and automatic adjustments for tone reproduction, gray balance, and color correction by Kodak\u27s PCS100 CMS were employed. The experimental design began with the calibration of each component in the imaging chain. Next, a three-dimensional test scene of objects displaying tone and color variety was digitized by the Lumina camera under specific studio lighting conditions. And, under the exact studio conditions, a Kodak Q-60 test target was digitized; this image file was used to characterize a device profile for the Lumina digital camera. The digitized 3-D test scene file was sent through a color-managed workflow for automatic color reproduction. IX The automated, color-managed reproduction process was as follows: 1) select monitor, input, effect, and output profiles in the PCS100 Color Manager 2) acquire image via Photoshop on a Macintosh 3) image color conversion with Kodak\u27s PCS100 plug-ins by applying custom input profile, output simulation profile, and 3M Matchprint output profile 4) film output via Agfa Selectset 5000, and 5) 3M Matchprint color proofing to SWOP (Specifications for Web Offset Printing). Subjective evaluation was based on the single stimulus method. Visual assess ments were performed by twenty color-tested judges with experience in printing or photography. A set of ten color proofs of identical image were individually evaluated for acceptability. The criteria for acceptable color reproduction includ ed tone reproduction, gray balance, and color correction. Proofs that received high average scores (\u3e80%) were determined acceptable. Analysis of the results determined that with proper calibration and CMS color conversion technology, one can deliver acceptable tone reproduction and pleasing color. Gray balance was determined unacceptable for all proofs based solely on a perceived yellowish-green cast in the MacBeth ColorChecker\u27s three-quarter tone patch. Excluding the gray balance factor, four proofs were determined acceptable for tone and color reproduction. Objective evaluation was made to further assess the color accuracy from original to acceptable proof, and to correlate colorimetric differences with the visual assessments. Quantitative assessment was based on colorimetric CIEL*a*b* mea surements and calculated color differences (AE, AL*, AC*, AHab*) of MacBeth color patches and 3-D objects. Objects in the original scene and corresponding image areas in the proofs were measured in order to study variations in hue, light ness, and saturation. Analysis of the results demonstrated that overall, the images in the proof were lighter, less saturated, and had small hue shifts compared to the original. The proofed image would probably be a poor match to the original in terms of objective color accuracy. But for this thesis project, color proof accept ability was determined by the subjective, visual evaluations