engleski

A STUDY OF SAMPLE CHOICE INFLUENCE ON THE ESTIMATED COLOR REPRODUCTION SYSTEM ACCURACY

The growing demands for quality and well defined product quality levels encourage the printing houses to align their production processes to the specifications defined by the international standards. Those standards prescribe the use of certain tools. As the tools, such as characterization and process control charts are standardized, they may be more or less suitable for certain color reproduction processes, as many studies have shown that the choice of training set values has a great impact on the characterization model accuracy. This study investigates the impact of the test set values choice on the estimated color reproduction system accuracy. Three random RGB value samples in each of the three different sample sizes have been taken, and their corresponding L*a*b* values predicted by the standard ICC profile were compared to the empirical values on the laser electrophotographic printer with RGB driver. The results obtained with different sample sizes and different random samples were compared using the analysis of variance to find whether the samples are significantly different.

CRS evaluation; test set size

nije evidentirano