Method for Grade Identification of Tobacco Based on Machine Vision

Abstract. Tobacco grading is the first step in the transfer of tobacco leaves from agricultural products to commodities and is key to determining the quality of tobacco. Manual grading is conventionally used for tobacco grading. However, it is time-consuming, expensive, and may require specialized labor. To overcome these limitations, a method for grade identification of tobacco leaves based on machine vision is proposed in this article. Based on a fuzzy pattern recognition algorithm, the tobacco leaf samples of the model set and prediction set could be classified by extracting appearance characteristics of the tobacco leaves. The identification system for tobacco leaves based on fuzzy pattern recognition was developed in MATLAB. The rate of correct grading was 85.81% and 80.23% for the modeling set and prediction set, respectively. This result shows that machine vision based automatic tobacco grading has a great advantage over manual grading, and this method can be explored for viable commercial use.