Synthetic reflectance curves by subtractive colour mixing

Abstract

An algorithm has been developed, using the subtractive primaries, that enables a ‘synthetic reflectance curve’ to be calculated for any set of XYZ values. A wide range of reflectance data for cyan, magenta and yellow primaries and white substrate can be used as data for the algorithm, provided the curves are smooth and correspond to those colours. These synthetic reflectance curves can be used as replacements for the data from measured physical standards for match prediction purposes. The synthetic standards have good colour constancy under a range of illuminants. They have been used to speed up the generation of fashion shade ranges selected initially on a CAD system.

Reconstruction of reflectance data by modification of Berns' Gaussian method

“…1,[4][5][6][7][8][9][10] In this section, the most customary methods i.e., principal component analysis (PCA) technique, Hawkyard's iterative as well as typical Berns' methods are briefly reviewed. Finally, the new methods for the recovery of reflectance curves which have been derived from Berns' method are presented.…”

“…Since the reflectance data of nonfluorescent materials are independent of the spectral power distribution of the illuminant and the spectral responses of observer, so it is considered as the ''fingerprint'' of a colored surface. 1 The reflectance data are necessary in many cases and the most important applications are:…”

“…For instance, the spectral data are inaccessible when a tristimulus colorimeter rather than a more expensive spectrophotometer is used. 1 Although the spectral data of a colored surface can be easily converted to the color coordinates, the procedure is not invertible and the estimation of reflectance data from the tristimulus values is not possible by simple mathematic routines. In fact, the multidimensional spectral space can be simply reduced into three-dimensional color space whereas, the conversion procedure is known as ''one-to-many problem'' and needs more mathematical knowledge.…”

“…Many mathematical techniques such as simplex method, genetic algorithm, simulated annealing, application of subtractive and additive primaries, neural networks, and principal component analysis technique have been introduced for extracting reflectance data from the relevant color coordinates. 1,[4][5][6][7][8][9][10] In this article, the Berns classical method for reconstruction of reflectance data, 6 which benefits from exceptional simplicity has been modified by using different types of primaries. Firstly, the Gaussian red primary has been changed to sigmoidal one for better simulation of the reflectance curve of natural objects in the red parts of visual spectrum.…”

Recovery of reflectance spectra from CIE tristimulus values using a progressive database selection technique

“…The spectral data are defined as a ‘fingerprint’ of an object and is very critical in many applications [1]. For example, it is essential for the estimation of the object under different illuminations, providing input of computer programs for the instrumental colour‐matching algorithm of textiles, paints, plastics, inks and computer graphic applications that require a wavelength‐based approach to specify colours [2].…”