Multiclass Classification of Tomato Leaf Diseases using a Hybrid of LSTM and CNN Model

Identifying diseases on tomato leaves is very important as tomatoes contain a rich amount of vitamins and other essential compounds for the human body. It helps keep crops healthy and makes farming last longer. Tomato leaf diseases are big problems for farmers and the farming business. Some of the key challenges associated with tomato leaf diseases include: Plant diseases in the leaves of tomato plants directly lead to a decrease in crop production. Diseases like early blight, late blight and bacterial wilt cause leaves to fall off too soon. This results in the plant's ability to grow fruit properly. Diseases in leaves can affect tomatoes by decreasing fruit production with weird shapes and marks. This not only makes the food look bad but can also result in people saying no to buying it and making its worth lower. In the proposed work, we proposed a strong ensemble model that uses Long Short-Term Memory (LSTM) and Convolutional Neural Network (CNN). This helps to accurately identify ten different types of tomato leaf diseases using a learning rate of 0.015, our hybrid model gets an accuracy at 99.8%. The model's performance is shown more by a loss value of 0.032, showing how good it is at finding complicated patterns in the data set. This smart method not only shows the strength of mixing LSTM and CNN designs, but it also helps in improving the classification accuracy of tomato leaf disease detection. It provides a trustworthy tool for farmers and scientists working on growing plants for many kinds of disease detection in tomato plants.