The effect of substance concentration on the spectral information of a multi-spectral sensor

Abstract

This research aimed to investigate the effect of substance concentration on the spectral information of a multi-spectral sensor. The measurement unit contained an Arduino UNO microcontroller that was used to read the spectral data obtained from a multi-spectral sensor. The multi-spectral sensor received reflected light from the samples, which were coffee phantoms with various coffee concentrations in the range of 0.5 - 60 mg/dl. Normalized spectral data were analyzed using analysis of variance and regression methods to investigate the effect of coffee concentration on the normalized spectral data of the multi-spectral sensor. According to the results of the study, the coffee concentration is related to the normalized data in the form of a polynomial function. The normalized spectral data inversely varied with the coffee concentration. The quadratic equation obtained from the regression at wavelengths of 580-600 nm was selected as the prediction model for measuring the coffee concentration. The predicted concentrations agreed with the actual concentrations. Therefore, spectral sensors can be applied to the concentration measurement of some colored biological substances.