Prediction of a taste from color and its practical application

Abstract

The relationship between the attributes of color and the perceived quality of taste was investigated. A survey with 100 Korean adults was conducted, in which 120 color patches were presented to the subjects. The subjects were asked to evaluate perceived taste in terms of sweetness, sourness, bitterness, saltiness, spiciness, and nuttiness. The multiple regression analysis was performed in order to derive formulae by taking the hue(a*, b*), saturation(C=√(a² + b²)), and lightness(L*) of the CIE 1976 L*a*b* values as the independent variables. It was revealed that the perceived taste intensity is affected by the combination of hue, saturation, or lightness at the significant level of 0.01. Consequently, it was possible to derive equations for each taste. Among the five tastes, bitterness was most predictable (R2= 0.24) followed by spiciness (R2= 0.23). As an extension to a practical application, a test with six types of food—chocolate, vitamin C, capsule coffee, spam, cup ramen, and cheese—was carried out (N= 50), and each type of food had four to six variations of color. The formulae were applied to rank the colors of each food in terms of its dominant taste and the result shows that the formulae generally have enough validity in predicting the taste intensity of the food through color.