Abstract

As plants are exposed to outer environment, they are highly prone to many diseases which effect the crop yield and results in low production. Due to lack of proper care of plants and regular plant disease check, severe effects are caused on plants. Agricultural productivity is something on which economy highly depends. Plant pathologists require a reliable and effective method to diagnose the disease effectively. Several methods and techniques have been applied to get correct disease classification. However, we need a method which has automatic identification, accurate result with minimum time. Previously many Machine Learning models have been tried and tested for the detection and classification of plant diseases but, after the much advancements in a subset of Machine Learning, i.e. Deep Learning, this area of research appears to have great potential as it fulfils the requirements with an increased accuracy. The AlexNet model is trained using an open dataset with 38 different classes of plant leaves and background images. The model was trained using different training epochs, batch sizes and dropouts. After an extensive simulation, the proposed model achieves 97.66% classification accuracy. This accuracy of the proposed work is greater than the accuracy of traditional machine learning approaches. The proposed model is also tested with respect to its consistency and reliability.