Tea Leaves Identification Based on Gray-Level Co-occurrence Matrix and K-Nearest Neighbors Algorithm.

In order to detect and identify tea varieties more efficiently, an identification project based on Gray-Level Co-Occurrence Matrix (GLCM) was developed. First, different kinds of typical tea which are Anji white tea, Biluochun, Dahongpao, Tieguanyin and Yixing black tea need to be collected and used as first-hand information. Then, five different types of Chinese tea images were taken as raw inputs, each of which had 25 images. The 125 images were processed through MATLAB software. After a 7-fold cross-validation step, the project showed a good result. The results showed that the matching degree of tea varieties reached 86.4%. This scheme provides an effective approach for tea identification process [ABSTRACT FROM AUTHOR]