Conjoint Analysis of Parametrized Gamut Mapping Algorithms

摘要

We show that conjoint analysis, a popular multi-attributernpreference assessment technique used in market research, is arnwell suited tool to simultaneously evaluate a multitude of gamutrnmapping algorithms with a psycho-visual testing load not muchrnhigher than in conventional psycho-visual tests of gamut mappingrnalgorithms. The gamut mapping algorithms that we test usingrnconjoint analysis are derived from a master algorithm by choosingrndifferent parameter settings. Simultaneously we can also testrnthe influence of additional parameters like gamut size on the perceivedrnquality of a mapping. Conjoint analysis allows us to quantifyrnthe contribution of every single parameter value to the perceivedrnquality.