ifferent reproduction devices can have different sets of re-producible colours. These sets are called gamuts. The processof transforming colours from one device (or image) gamut to an-other is called gamut mapping. Gamut mapping has many techni-cal issues to be considered: the used colour space, direction andmagnitude of the mapping and whether and to which extent in-gamut colours should be altered. Spatially invariant algorithmstreat all the pixels independently on their position in the image.Spatially variant (local) algorithms allows a better rendition butintroduces the problem of artefacts and/or haloing in the result-ing image. In this paper we propose a spatially variant gamutmapping algorithm that creates virtually no artefacts nor haloingin the resulting image. We start from an analysis of the Retinexalgorithm and devise proper functionals to build an algorithmwhich tries to maintain spatial ratios in the image while map-ping it into the gamut and, at the same time, avoids all drawbacksof Retinex approaches. We suggest to perform the mapping in anRGB colour space rather than one of the perceptually more ho-mogeneous ones. Although less homogeneous, we experimentallyshow that RGB colour spaces actually have better hue constancyaccording to a certain criterion.
展开▼
