Proposed CNN Model for Tea Leaf Disease Classification
Rahul Singh, Neha Sharma, Rupesh Gupta
Abstract
Tea leaf diseases have had a substantial impact both on the quantity and quality of the tea produced. The high-precision automatic detection and identification of illnesses that can be found in tea leaves is beneficial to the accurate prevention and control of those diseases. Manual procedures, which require a lot of time and effort, are still the primary tool for diagnosing tea illnesses and determining the severity of their effects and also effect the agriculture. This situation has persisted for some time. It is helpful to the tea leaf disease prevention and control efforts to have accurate and speedy disease detection. This research presents a technique for tea leaf disease classification that is based on an improved version of a deep convolutional neural network. This project aims to develop a deep convolutional neural network model capable of identifying diseases affecting tea plants based on image sets of their leaves. The results of the studies indicate that the proposed method has an average identification accuracy of 73%, which is higher than the accuracy of more conventional manual approaches. In later applications, the CNN model was utilized to improve the diagnostic measurement of tea leaves as well as the measurement of leaves from other plants.