Detection of Disease in Tea Leaves Using Convolution Neural Network
Shyamtanu Bhowmik, Anjan Kumar Talukdar, Kandarpa Kumar Sarma
Abstract
Tea leaf diseases such as Black Rot and Rust of Tea are a major critical threat to Indian food security. The reason being is rapid identification of these diseases still remains difficult in many corners of the world till today because it requires huge amount of work, prior knowledge in the concerned diseases, and also requires the huge processing time. Diseases are always injurious to the tea plant's health which in turn severely affect its growth. In 2018, tea industry contributes almost 6% in the Indian economy. So it is always advisable to continuously monitor its growth to ensure minimal losses of this tea plant. But it is not possible to monitor this using naked eye. So this paper basically adopts convolutional neural network (CNN) model to detect and identify these two disease using convolutional neural network (CNN). The proposed works basically finds a solution of the tea leaf disease detection using simplest method while keeping minimum computational complexity and minimal resource to gain fast and accurate result as convolutional neural network (CNN) automatically extracts features for classification of input image into various classes. The experimental results on the developed model achieved precision approximately 95.93%.