DDTNet: A Novel Approach to Detect Disease in Rice Leaf using Deep Learning

Fifty percent of world’s population depends on rice, hence its production is crucial to guarantee the stability of food supply all around. Rice crop diseases may result in considerable output losses, which can affect both the availability of food and the lives of farmers. Early diagnosis and identification of these disorders are necessary for appropriate intervention and care to prevent further complications. This work introduces a deep learning-based method for disease identification in rice leaves. The system makes use of the capabilities of convolutional neural networks (CNNs) to effectively categorize various kinds of diseases. We proposed DDTNet (disease detection network) to detect disease in rice leaves. The proposed model has 4 conv2D layers along with max-pooling and output layer to correctly diagnose the disease. Through training and validation on a dataset consisting photos of rice leaves, the DDTNet achieved an F1 score of 95.5%, accuracy of 95.6%, and precision of 94.8%. The model has been developed to illustrate its potential as a tool for precision agriculture. Within the context of real-time applications, the model may be implemented on the cloud.