A New Method for Color Gamut Mapping by Genetic Algorithm

T o reproduce an image, it is ‎necessary to map out of gamut ‎colors of the image to destination ‎gamut. It is clear that the best ‎color gamut mapping introduces ‎the perceptually closest image to ‎the original one. In this study, a ‎new color gamut mapping is ‎purposed by the aid of Genetic ‎Algorithm (GA). The color ‎difference between the original ‎and mapped images based on S-‎LAB formula was chosen as fitness ‎function. The proposed algorithm ‎was applied in CIELAB color ‎space and special genetic ‎operators were developed to meet ‎the aim of gamut mapping. To ‎increase the rate of convergence ‎and have a faster algorithm, one of ‎the initial population ‎chromosomes can be obtained ‎from the result of clipping method. ‎The results showed that the new ‎method introduces smaller color ‎difference between the reproduced ‎and original images in comparison ‎with the common clipping method. ‎The other advantage of the genetic ‎color gamut mapping is that any ‎new criterion for color image ‎difference can be easily used as a ‎fitness function. In addition, by this ‎method the final colors are not ‎restricted to the gamut surface and ‎they may be included into the ‎gamut. ‎