A Multi-Ink Color-Separation Algorithm Maximizing Color .

original target(filled square) and reproduction by CMYKcm (filled triangle) andCMYKGO (filled circle) printers. Samples 1 – 20 correspond to theKodak Gray Scale; samples 21 – 44, the ColorChecker.The reflectance spectra from the Kodak Gray Scale #5and the Orange Yellow Color Checker samples are plottedin Figure 6. Their CII values are listed in Table I. For theneutrals, it is clear that the most spectrally flat sample hadthe best color constancy. In fact, only object colors withthese properties have guaranteed color constancy. TheCMYKcm profile did not use black, hence the large spectralselectivity and meagre color constancy. When analyzing theOrange Yellow, predicting which sample has the best colorconstancy based on spectra is not as straightforward. Themore compound spectrum of the CMKYGO sample hadbetter color constancy than the original ColorCheckersample. As shown by Berns, et al., the location of theabsorption peaks may be the key factor rather than theirdegree of spectral selectivity. despite of the underlyingcause, it is clear that the CMYKGO print had enhancedcolor constancy compared with both the CMYKcm printand many of the chromatic samples of the ColorChecker.Figure 6. Reflectance spectra of original and printedreproductions using CMYKGO and CMYKcm ink sets for KodakGray Scale Sample 5 (top) and Orange Yellow ColorCheckersample (bottom).ConclusionsAn algorithm was developed to deal with the one-to-manymapping problem when building color lookup tables formulti-ink printing. The unique feature of the algorithm isthat maximal color dependability was the main selectionIJSER 2013http://www.ijser.org

International Journal of Scientific & Engineering Research, Volume 4, Issue 12, December-2013ISSN 2229-5518criterion among ink combinations yielding similar color.The algorithm was tested with spectral data from numeroustargets. The algorithm as implemented can be improved by abetter CIELAB allotment of the virtual sample set.Furthermore, one can imagine a number of criteria that canbe used as a selection metric, individually or combined,such as graininess, sharpness, maximum black ink amount,print precision, etc. Future research will focus on tradeoffsbetween these various criteria in terms of print quality. Atheoretical analysis is also justified to understand the interrelationships between the number of inks, their spectralproperties, color constancy, and color gamut.References[1] V. A. Bardadym, “ Computer-Aided School and UniversityTimetabling: The New Wave,” in the Practice and Theory ofAutomated Timetabling: Selected Papers from the 1st InternationalConference (Lecture Notes in Computer Science 1153), E. Burkeand P. Ross, Eds. Berlin, Germany: Springer-Verlag , pp. 22–45,1996.[2] Burairah Hussin, Abd Samad Hasan Basari, and et.al , “ExamTimetabling using graph colouring approach”, 2011 IEEEConference on Open Systems (ICOS2011),pp. 133-138, 2011.[3] Gend Lal Prajapati1, Amit Mittal, “An Efficient Colouring ofGraphs Using Less Number of Colours”, IEEE , pp. 666-669 ,2012.[4] D. Werra de, “The combinatorics of timetabling,” EuropeanJournal of Operational Research, 96 (3) pp. 504 –513, 1997.[5] P. Cheeseman, B. Kenefsky, and W. Taylor, Where the reallyhard problems are, In J. Mylopoulos and R. Reiter (Eds.),Proceedings of 12th International Joint Conference on AI (IJCAI91), Volume 1, pp. 331–337., 1991.[6] D. Brelaz, “New methods to colour vertices of a graph,”Communications of the ACM 22, pp. 251 –256, 1979.[7] P. Galinier and J-K. Hao, “Hybrid evolutionary algorithms forgraph colouring,” Journal of Combinatorics, 3(4), pp. 379 –397,1999.2231[8] Amit Mittal, Parash Jain, Srushti Mathur, Preksha Bhatt,“Graph Colouring with Minimun Colours” , 2011 InternationalConference on Communication Systems and NetworkTechnologies , pp. 638-641 , 2011.[9] Hertz, A. and D. de Werra (1987). Using tabu search techniquesfor graph coloring. Computing 39(4), 345-351.[10] Br′elaz, D. (1979). New methods to color vertices of a graph.Communications of the ACM 22, 251-256.[11] Turner, J. S. (1988). Almost all k-colorable graphs are easy tocolor. Journal of Algorithms 9(1), pp. 63-82.[12] Johnson, D. S., C. R. Aragon, L. A. McGeoch, and C.Schevon (1991, May-June). Optimization by simulated annealing:An experimental evaluation; part II, graph coloring and numberpartitioning. Operational Research 39(3), pp. 378-406.[13] Kauther Drira, Lyes Dekar, Hemamache Kheddouci, “A SelfStabilizing (delta 1)- Edge-Coloring Algorithm of ArbitraryGraphs”, International Conference on Parallel and DistributedComputing, Applications and Technologies, PDCAT, pp. 312317, 2009.[14] R. Qu, E. K. Burke, B. McCollum, L. T. G. Merlot, and S. Y.Lee, "A Survey of Search Methodologies and AutomatedApproaches for Examination Timetabling," The University ofNottingham, Nottingham 2006.[15] R. Lewis, "Metaheuristics for University CourseTimetabling," School of Computing. vo!. PhD thesis Edinburgh:Napier University, 2006.[16] A. Hertz and D. de Werra, “Using tabu search techniques forgraph colouring,” Computing 39 (4), pp. 345–351, 1987.[17] J. Randall-Brown, “Chromatic scheduling and the chromaticnumber problems,” Management Science 19(4), Part l, pp. 456 463, 1972.[18] N. Robertson, D. Sanders, P. Seymour and R.Thomas,"Thefour color theorem", Journal of Combinatorial Theory Series B 70:2–44, 1997.[19] M. Stone, "Color in Information Display," in Vis06, 2006.[20] H. Asmuni, E. K. Burke, and 1. M. Garibaldi, "FuzzyMultiple Heuristic Ordering for Course Timetabling." Proc of the5th UK Workshop on Comput Intel , pp. 302-309,2005.IJSERIJSER 2013http://www.ijser.org

ColorChecker Color Rendition Chart and a Kodak Gray Scale were used to create a CIELAB TIFF image for D50 and the 2 observer. The image was processed through the color-separation algorithm and printed on the CMYKGO inkjet printer. The image was also printed on an identical model printer with the standard CMYKcm ink set using the