Crop Disease Detection Using Deep Neural Networks

Abstract

Agriculture has benefited greatly from technological advancements. Farmers can readily connect with their farms using technology. Because of technology, all farming processes have improved. This paper uses Deep Learning Models and Machine Learning techniques, particularly the ResNet50, to identify pest attacks and diseases from leaf images of plants and accurately provide countermeasures in English and the local language, aptly enunciating the genus and scientific name of the pest, climatic conditions in which they thrive, appropriate countermeasures and type of pesticide application, as well as duration and time interval, followed by local language translation. The model can be replicated to other crops with suitable customization to achieve high accuracy in prediction. Overall prediction accuracy currently stands at 99.05% for Tomato crop and 99.52% for Potato crop.