Quantitative Determination of Colored Oily Soil Adhering to Metal Surface Using Digital Image Data

Abstract

A method for calculating the amount of colored soil on a flat metal surface from digital image data was examined. Round samples cut out of SUS tape were soiled with oily soil mixed with sudan III, placed in sample bottles, and washed using a tube rotator. Images of the samples before and after the washing process were captured using a CCD camera and the image data were processed by a computer. The shine from the metallic surface was controlled by using indirect lighting. It was necessary to diminish the effect of the substrate's color, and to this end, we attempted to apply linear and non-linear color correction procedures. We found that the use of gamma correction after linear correction to remove the effect of the metal surface color was the most accurate quantitative method. Additionally, we conducted a washing test with commercial detergents using this quantitative method for image data and found that the removal process for the greasy soil from the metal surface could be expressed as a first-order reaction.