Development of a colorimetric sensor array with weighted RGB strategy for quality differentiation of Anji white tea

Abstract

This study presents the development of an innovative colorimetric sensor array system designed to differentiate the quality of Anji white tea by detecting and analyzing its aromatic components. The colorimetric sensor array utilizes chemically reactive dyes that selectively respond to specific aromatic compounds, effectively capturing aroma variations across different grades of tea. Under optimized experimental conditions, response data from the colorimetric sensor array for various tea samples were collected and processed using a weighted RGB strategy, which assigns values of 30%, 59%, and 11% to the red, green, and blue channels, respectively, to better replicate human visual perception. Tea grade classification was performed using principal component analysis and orthogonal partial least squares discriminant analysis. Comparative analysis of the performance of the four supervised learning models demonstrated that the weighted RGB method was more accurate, with the Extreme Learning Machine achieving an impressive classification accuracy of 96.30%. This novel method provides a rapid, efficient, and user-friendly tool for tea quality assessment, with significant implications for the food engineering industry.