On the use of Computer Vision to Estimate Chemical Concentration based on Colorimetric Analysis

In this paper, we report an investigation on estimation of chemical concentration by using computer vision-based image processing techniques. Eleven liquid samples have been prepared by mixing copper (II) sulphate salt with distilled water with their concentrations varying from 0.0 - 1.0M (molar concentration). Images of the colored chemical solutions, kept in transparent test tubes, are acquired by using a camera module interfaced with Raspberry Pi single board computer. A customized python program has been developed by incorporating OpenCV libraries in order to analyzed the acquired images of the samples in LAB color space. Experimental observations have shown good correlations between the color component values and molar concentrations of the liquid samples. Novelty of the proposed method includes portability, rapid analysis, low-cost and applicability in out-of-the-lab scenario.