Page 195 - Computational Colour Science Using MATLAB

182 MULTISPECTRAL IMAGING
where, the matrix XYZ is an n63 matrix of tristimulus values and obs is a string
that defines the illuminant/observer combination (e.g. ‘d65_64’). A matrix P is
returned with the reflectance values with size 316n.

Box 26: xyz2r.m

function [P] = xyz2r(XYZ, obs)

% function [P] = xyz2r(XYZ, obs)
% estimates reflectance P from XYZ trimulus values
% matrix XYZ must be n by 3
% illuminants A, C, D50, D55, D65, D75, F2, F7, F9
% set obs to ’d65___64 for D65 and 1964, for example
% matrix P is returned as an n by 31 matrix

% check dimensions of XYZ
dim = size(XYZ);
if (dim(2) *=3)
disp(’XYZ must be n by 3’);
return;
end

load evectors.mat
% loads the 31 by 3 matrix v containing three basis
functions

load weights % contains the tables of weights
if strcmp(’a___64’,obs)
cie = a___64;
elseif strcmp(’a___31’, obs)
cie = a___31;
elseif strcmp(’c___64’, obs)
cie = a___64;
elseif strcmp(’c___31’, obs)
cie = c___31;
elseif strcmp(’d50___64’, obs)
cie = d50___64;
elseif strcmp(’d___50’, obs)
cie = d___50;
elseif strcmp(’d55___64’, obs)
cie = d55___64;