A Predictive Deep Convolutional Neural Network Approach on Agriculture Datasets for Tomato Leaf Disease Detection

In the world the most common crop, the tomato may be found in every kitchen in many ways, whatever of the type cooking. In the world, it is the third-most extensively grown crop, after sweet potatoes and potatoes. India came in second place for tomato output. However, the various diseases reduce the quality and output of the tomato crop. Hence, a deep learning-based policy for disease prediction is discussed in this research. To recognize and classify diseases, a technique based on Deep Convolution Neural Networks is used. This study proposes an AlexNet model to forecast nine different diseases in tomato leaves. For experimental learning, the freely accessible dataset known as Plant Village is employed. The suggested model shows a training accuracy of 97.73% with a stochastic gradient descent optimizer (SGD), having a batch size of 32 and 400 iterations. The total time taken to fit the module is 563 min overall. The experimental outcomes illustrate that the suggested method is successful in detecting tomato leaf disease and that it may be expanded to detect additional plant diseases. From the comparative analysis and experimental study with the other cutting-edge techniques, it may be decided that the suggested model has relevance to quickly identifying plant diseases.