PLANT DISEASE PREDICTION USING CONVOLUTIONAL NEURAL NETWORK

Abstract

Identification of plant diseases at an early stage is critical to forever maintaining agricultural productivity as well as ensuring food security. Often times relying on a manual inspection, traditional methods of diagnostics can be slow and inaccurate as well. This research proposes a new automated method with the use of Convolution Neural Networks (CNNs) to diagnose and classify plant diseases with the aid of their leaves images. This model uses a vast dataset consisting of different crops as well as different diseases. The model that is being proposed, CNN, is capable of feature extraction making it easier for the model to recognize the disease. After sufficient training and validation, the model can distinguish and tell with great precision whether the plant is healthy or infested with some form of disease. This technology can easily be integrated in mobile systems so that it can be made available to farmers and this can help them in real time diagnosis which enables them to make timely decisions and prevent losses. This study shows how Deep Learning can be evolutionary for agriculture with available and effective measures for managing the health of plants.