A digital image processing in combination with a personal computer and a CCD camera was applied to the evaluation of the amount of soil for cleaning studies. Four types of substrate/soil model i.e., (1) glass/oily soil, (2) PET film/sugar soil, (3) cloth/oily soil, and (4) carpet/particle soil models, were used for experiments. The components of the oily soil were beef tallow, soy bean oil, mono-olefin and oily colorant. The sugar soil was mixed with Food Red No.104 as a tracer. A carpet was soiled by Sudan III in solid state by rubbing. Three of these models, (1) - (3), were washed in water. RGB values from digital images were transformed into reflectance. Kubelka-hunk equation was applied to the calculation of the amount of soil from the digital image with a PC value.
Good correlations were found between the PC values and the amounts of the soil calculated from the absorbency of the extracted solution for all substrate/soil models. Coefficients of correlation in this study range 0.9299 to 0.9876. These results suggest that this system would be useful in practical use.
View full abstract