COLOR CORRECTION USING PRINCIPAL COMPONENTS

The problem of color correction for changes in illumination is briefly reviewed. A new method using the spectral reflectance distribution in the image is introduced and compared to a white-point mapping method. The new method is based on principal component analysis and thus is dependent upon the characteristics of the images that are to be viewed. These two methods are compared using chromaticity plots and the CIE L*a*b* DELTA-E measurement, for test samples which include color copiers, thermal writers, ink-jet printers, and lithography.