Quantifying Nonhomogeneous Colors in Agricultural Materials Part I: Method Development

Abstract

Measuring the color of food and agricultural materials using machine vision (MV) has advantages not available by other measurement methods such as subjective tests or use of color meters. The perception of consumers may be affected by the nonuniformity of colors. For relatively uniform colors, average color values similar to those given by color meters can be obtained by MV. For nonuniform colors, various image analysis methods (color blocks, contours, and "color change index"[CCI]) can be applied to images obtained by MV. The degree of nonuniformity can be quantified, depending on the level of detail desired. In this article, the development of the CCI concept is presented. For images with a wide range of hue values, the color blocks method quantifies well the nonhomogeneity of colors. For images with a narrow hue range, the CCI method is a better indicator of color nonhomogeneity.