Prediction of the best-fit regression model to correlate instrumental colour measurement and visual assessment

The results of a study comparing instrumental colour measurement and visual assessment are reported. The study made use of a comprehensive set of visu al colour-difference pairs including 240 dyed samples distributed around 20 colour standards and spread evenly throughout colour space. Regression ana lysis was then used to attempt to describe the relationship between the vis ual and instrumental results. This showed that linear and logistical regres sion gave the two best-fit models. This work also provided a useful data se t for evaluation of colour-difference equations so that the optimum colour tolerance levels of various colour data sets could be determined.