NEURAL NETWORKS MODEL ON COLOR RECIPE PREDICTION

Yıl 2011, Cilt: 18 Sayı: 81, 12 - 21, 01.01.2011

Nurhan Onar

Öz

Color recipe prediction was known to predict which colorants and their concentrations were required for a given reference color. Making a good color recipe prediction is even hard to expen colorists. Color recipe prediction can be made by classical and computering recipe prediction systems. The classical system is based on a colorist's prediction, while the latter is applied as linear model (Kubelka-Munk model) in common or nonlinear mode! (such as neural network model). In this paper, how color recipe prediction by neural network based on nonlinear model is working was reviewed. 

Anahtar Kelimeler

Colorre<:ipc prediction, K-M mode!, neural networks

RENK REÇETESİ TAHMİNLENMESİNDE YAPAY SİNİR AĞLARININ KULLANIMI

Öz

Rcnk rc"ctc tahmin1cmcsi; veri len biT rderans rcnk ilc aym rcngi urctmck iyin hangi boyamladdclcrin hangi konsantrasyonlarma ihtiya~ duyuldugunu tahmin ctmcktir. iyi biT tahminlcmc yapabilmck dcncyimli boya uzmanlan i"in bile zordur. R~C1C tahmini; klasik rc\,etc ~lkarma sistcminc gore veya bilgisayarh rc,"c\c !,'lkam13 sistcminc gore yapllabilir. Klasik rc,"cic vlkanna sistcmindc biT boya uzmalll dcncyimlcrinc dayanarak vcrilcn renge yaklll biT r~ctc tuttunnaya (j:ah~lr, bilgisayarh rcvctc (j:lkarma sistcmlcri isc yaygm olarak bilincn dogrusal modele (Kubclka-Munk modcli) veya dogrusal olmayan (yapay sinir aglafl) modele dayanabilir. Bu makalede dogrusal olmayan sisteme dayanan yapay sinir aglarmm kullanl ml ile reo;ete O;lkanna sisteminin nasi! ~

Anahtar Kelimeler

Renk r~etesi tahmini, K-M mode!i, yapay sinir aglan