Using Appearance Maps Drawn from Goniocolorimetric Profiles to Predict Sensory Appreciation of Red and Blue Paints

摘要

In this article,we test the ability of a sensorialy calibrated version of Appearance Maps,a recently suggested tool based on Multidimensional Scaling and Procrustes Analysis,to understand and predict the sensory evaluation of automotive paints by trained subjects.This tool has been applied to a set of 13 industrial samples representing three shades: two effect shades,both at 4 levels of quality,and one solid shade at 5 levels of quality,which were evaluated for their overall appearance relative to the distinctness of image and the"Orange Peel"effect.The results of descriptive analysis performed on the samples by a panel of 15 trained panelists(paint experts and sensory assessors)are compared with information taken from the maps.One of the Appearance Map coordinates is shown to correlate well with mean sensory evaluations.Different relations(either sigmoid curves or power laws)between map coordinate considered as a stimulus and sensory scores taken as perception are found for the different shades,and comments on possible similarities between the observed relations and known models are made.