Oryza Sativa: Image Classification Using Deep Learning

Rice being a staple food source for a huge population of the world, it is very crucial for their health and productivity in terms of food security. However, the rice plants suffer from various diseases which significantly reduce crop yield. A CNN-based model for the automated classification of common rice leaf diseases, namely Bacterial Leaf Blight, Brown Leaf Spot, Blast, and Sheath Blight, is introduced. The model is trained on a labeled set of rice leaf images, categorized into specific disease categories. The approach includes techniques of image preprocessing and augmentation for enhanced robustness in detection and classification. Experimental outcomes show that the proposed model can achieve near-perfect detection accuracy and outperform the existing methods with respect to determination of whether the leaf is healthy. This automatic system is supporting farmers and agronomists in managing crop health, hence reducing the disease impact on rice production.