Leaf Disease Detection Using Deep Learning

Agricultural productivity plays an important role in India’s economy, contributing to food and cash crops crucial for the environment and human well-being. Plant diseases can considerably impact the value and production of agricultural products. Inadequate disease diagnosis and the lack of knowledge about symptoms and treatments result in the death of many plants. The automatic plant disease detection field is gaining prominence, offering advantages in monitoring large crop fields and identifying disease symptoms on leaves. This paper focuses on detecting plant diseases to reduce crop losses and enhance production efficiency. Deep learning is used to detect plant diseases, and a Convolutional Neural Network is used to classify leaf images into 39 different categories. The model is trained on a dataset consisting of cases. To expand its size, we employed six distinct augmentation techniques. The achieved test accuracy is 98.9%. Various performance metrics are derived to evaluate the model’s performance.