Obtaining Linear Equations by Regression Method for Color Management
In this study, cyan, magenta, yellow and black colors prints were produced in 4 different digital printing machines
and the results were determined with color measuring device and transferred to computer. The main aim is to approximate
these printed colors to their original reference values and thus reducing the color variations caused by the differences between
of the printing machines. In this study, the L*a*b* values were used as the color space of the prints obtained from the printing
machines. This color space is geometric arrangement in visible areas (L*a*b*). A defined color space model provides an
objective classification criterion so that colors can be expressed numerically. Linear regression method was used to
approximate the print color values to the original value. Here, the coefficients of the second order equations were determined.
The Gravitational Search Algorithm (GSA) method was selected to determine the coefficients of the equation. This method
calculates the coefficients of equations so that the difference between the L * a * b * values of the colors obtained from the
printing machines and the L * a * b * values of the original colors is minimized. All process was performed in MATLAB
software. The obtained results show that the different printing presses, which were different from their original values, are
approximated to their original values thanks to the defined equations. These equations can be used in different printing
machines for producing closer printing results to original colors.
Index Terms - Color management, linear regression method, Gravitational Search Algorithm.