Classification of Tea Leaf Diseases Using Convolutional Neural Network

Abstract

Assam is the highest tea-producing state in India. The economy of this state is greatly dependent on the cultivation and productivity of tea. The biggest challenge to tea growers is to produce tea without microbial or pesticide damage. Since leaves are the harvested product in tea, so leaf diseases play an important role. Lack of awareness and care to leaves causes unfavorable impacts on plants, product quality, and quantity get reduced. The symptoms of the disease can be observed on the leaves. The leaf shows symptoms by changing color or showing spots on it. The identification of these diseases is made manual, which can consume more time or may be costly. The idea is to identify and classify the diseases accurately from leaf images automatically. The Convolutional Neural Network is being proposed in this study, which has classified the diseases with an accuracy of 92.59% and is more accurate than the prevailing classifiers like Support Vector Machine and K-Nearest Neighbors for this specific purpose.