A New Method for Color Gamut Mapping by Genetic Algorithm

To 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 SLAB 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. Prog. Color Colorants Coat. 2(2009), Institute for Color Science and Technology.