Quantifying the thickness of each color material in multilayer transparent specimen based on transmission image

Abstract

Combing the color transmission image and the Beer–Lambert law shows a great application prospect in quantifying each material in multilayer specimen. Here, a novel, low-cost, and efficient optical algorithm is proposed to predict the thickness of each color material in a multilayer specimen from the color transmission image based on the Beer–Lambert Law. In this work, a normal scanner is employed to achieve the color transmission image of the monochrome transparent films. RGB values represent the transmitted intensity. A linear relationship between the optical depth and physical thickness is observed under different monochromatic lights. It is supposed that for a multilayer transparent film which consisted of different monochrome transparent films, the optical depth is related to the physical thickness of each monochrome transparent component. Therefore, an estimating equation is proposed to predict the thickness of each color material in the multilayer specimen. According to the result, the standard deviation of predicted thickness and practical thickness of each color film in the multilayer specimen is 0.93%. Fairly good agreement and high accuracy are obtained between the practical and predicted values, and the validity of this method is confirmed.