In this study, a scanner has been used to compare the different characterisation methods, iteration and regression, with and without linear RGB and with using either XYZ or L*a*b*. The results show that the iteration method which marginally improves the characterisation accuracy with using XYZ values, has no significant effect in characterisation with L*a*b* values except for the first order polynomial. Also, it was shown that for the regression method either of the L*a*b* or XYZ approaches with an appropriate linearisation method can be used. However, if there is no technological problem, the L*a*b* approach is preferable due to having an optimisation criterion which directly relates to the visual perception of colour difference.
H. Izadan, J.H. Nobbs, "Input Device Characterisation: A Comparison between Iteration and Regression Methods Using either XYZ or L*a*b*" in Proc. IS&T CGIV 2006 3rd European Conf. on Colour in Graphics, Imaging, and Vision, 2006, pp 158 - 162, https://doi.org/10.2352/CGIV.2006.3.1.art00031