Electronic Tongue for Tea Quality Assessment

Abstract

Biochemical means of tea quality evaluation is quite accurate, but it requires very costly instruments and takes a long time to conduct the experiment. Researches show that thearubigin (TR) and theaflavin (TF) are the two most important biochemical compounds present in tea liquor on which liquor characteristics of black CTC tea depends. Consequently, this fact may be the basis of determining tea quality by assessment of TR and TF via electronic tongue (ET) response. This technique is free from subjective factors. In literature, there are only two research works using this technique so more research work is required to address the problem. This chapter proposes a modeling technique of ET response using sparse decomposition technique to estimate TR and TF content in a given tea sample. For each tea sample, sparse model coefficients obtained from the ET response is considered as a characteristic attribute of it. Experimental results using dissimilar regression models show high prediction accuracies which justify the efficacy of the proposed method.