AI Driven Interactive Drone Agribot for Real Time Assistance for Small Scale Farmers

The primary goal of this work is to help farmers and people identify plant diseases more precisely or to confirm if they have been afflicted by the illness or not. Our proposed model was trained from scratch using a set of plant leaf picture samples to categorize and identify the presence of illnesses. In order to extract characteristics from a particular plant leaf image and categorize it to determine whether any of the plants are sick with various diseases, we typically start from scratch and build a convolutional neural network model. This paradigm could help lessen the difficulties in interpretability and reliability that arise while working with agri-imagery. For this classification work, it is difficult to obtain a substantial amount of plant disease leaf dataset, in contrast to other deep learning classification problems where there is a sufficient image repository. As a result, we frequently use a variety of information augmentation techniques to improve the CNN model's validation and classification accuracy, and we frequently attain impressive validation accuracy. The 96% accuracy is attained by our trained model.