Characterization of tea (Camellia sinensis) granules for quality grading using computer vision system

Abstract

Tea (Camellia sinensis) has been found as an important medicinal beverage for human which is consumed all over the world. Primarily, the majority of tea is being cultivated in Asia and Africa, however it is commercially produced by more than 60 countries. Though substantial amount is produced, its processing system is still underdeveloped which leads to decrease in export opportunity as well as low monetary value. Moreover, the traditional method of tea grading and sorting is laborious, inefficient, and costly which ultimately produces the low-quality heterogeneous products. Processing and grading of tea granules after drying is very important task for maintaining quality. Computer vision (CV) applications in processing unit especially in grading and sorting of agro-products is very popular and reliable option to improve quality of produce. In this study, an attempt was taken to develop a machine vision system for quality grading of tea granules based on physical parameters of four standard tea grades namely BOP, GBOP, CD and PF. An image acquisition system with suitable illumination arrangement was developed to obtain high resolution image of tea granules. The images were analyzed to extract physical features like projected area, circularity, roundness, ferret diameter, aspect ratio and solidity. Tea granules (BOP, CD, PF and GBOP grade) were found significantly different for the textural features area, perimeter, circularity, roundness and ferret diameter. Projected area, perimeter, and feret diameter treated as a good indicator of the extracted features as the system has been able to significantly (p < 0.01) differentiate among the grade of tea. The developed characterization attributes based on physical features prior to an automatic sorting technology will improve the efficiency and enhance the cost-effectiveness which ultimately led to energize the international export market.

Keywords

• Tea
• Classification
• Computer vision
• Image processing
• Grading