ABSTRACT
The physical characteristics of made black crushing, tearing, and curling (CTC) tea were investigated by machine vision, and a quality grade estimation technique was developed by the application of neuro-fuzzy algorithms. The physical characteristics such as average size, shape (aspect ratio), and surface texture of the tea granules along with infused liquor color (after tea brewing) were acquired using an image acquisition system consisting of a machine vision setup. Data analysis in the form of principal component analysis (PCA) was performed to extract the significant physical attributes. The results of PCA were further investigated to obtain the correlations among tea samples and quality grades. The most significant physical attributes were fed to intelligent grade estimation software based upon neuro-fuzzy algorithms developed in MATLAB® software by using an adaptive neuro-fuzzy inference system (ANFIS). The proposed technique showed an accuracy of about 95%. The precision and repeatability by this technique was found be closer to the tea tasters’ perception, which can impart assistance to the tea tasters in a user-friendly and rapid manner, leaving aside the cumbersome chemical analysis.
