* Corresponding author. Tel.: #1 512 245 3189; fax: #1 512 245 3089; e-mail: ct01@swt.edu. Int. J. Production Economics 58 (1999) 31 — 37 A fuzzy decision model for color reproduction Cecilia Temponi*, Farid Derisavi Fard, H.W. Corley  Southwest Texas State University, San Marcos, TX 78666, USA  Xerox Corporation, Dallas, TX, USA  The University of Texas at Arlington, Arlington, TX, USA Received 4 June 1997; accepted 29 December 1997 Abstract Color quality appraisal refers to a visual evaluation of an original color image and an assessment of the degree to which that image is matched in photomechanical or electronic reproduction. The evaluation of reproduction quality is quite subjective since it relies on the individual judgments of color technicians who have different perceptions of color and its component qualities. To reduce the amount of rework due to subjective evaluation of color reproduction, an objective model is proposed here. In this model, a fuzzy set of quality factors is introduced and a numerical quality value calculated. This measure is then compared with an acceptable quality value experimentally established for particular customers and printing processes. Finally, the installed model is tested in actual production, and compared with the conventional evaluation method in a cost analysis.  1999 Elsevier Science B.V. All rights reserved. Keywords: Printing process; Color appraisal; Production; Fuzzy model; Cost analysis 1. Introduction In color quality appraisal, the original color image (color transparency, color print, or other reflection copy [1]) and the color reproduction are compared for hue, lightness, and saturation. They are also appraised in terms of tonal range, contrast, and texture, which affect the overall appearance of the color image. However, such analysis is subjec- tive and varies from one technician to another. This fact is the primary cause of production rework in the prepress industry today when the color repro- duction does not meet the customer’s expectation. Quantitative decision techniques have not been heretofore used, but the situation seems a natural application of fuzzy logic. In this paper we construct a model for the nu- merical evaluation of the quality level of a color reproduction by introducing a set of factors for a color technician to use in their evaluation and a measure of their respective importance. This ap- proach enables the technician to judge the degree of color match acceptable to a particular customer for a specified printing process. 0925-5273/99/$ — see front matter  1999 Elsevier Science B.V. All rights reserved PII: S 0 9 2 5 - 5 2 7 3 ( 9 8 ) 0 0 0 7 8 - 4