A visible colorimetric sensor array based on chemo-responsive dyes and chemometric algorithms for real-time potato quality monitoring systems

Abstract

The colorimetric sensor array real-time monitoring system with multivariate analysis was established for discrimination of potato varieties with different types and degrees of corruption. The characteristic volatile compounds of fresh, dry rot and soft rot potatoes was identified by Gas Chromatography-Mass Spectrometry and the 3 × 4 array was fabricated to capture the characteristics volatile compounds. The sensor array system produced a visible color difference map upon its exposure to volatile compounds of potato. Discrimination of potatoes with the same types or different degrees of corruption was subsequently achieved using principal component analysis and hierarchical clustering analysis dendrogram. The k-nearest neighbor algorithm for potato classification provided the best results with 100 % discrimination on both the calibration and prediction sets. The linear discriminant analysis model achieved a 99.76 % calibration set and a 99.31 % prediction set for potato grading. An online warning device based on array was devised to realize unmanned monitoring for potato quality.