Deep Convolutional Neural Network (DEEP-CNN) for Multi-Class Classification of Biotic Stress in Paddy Crop

Rice, as a staple food for billions, faces severe yield threats from biotic stress factors such as pathogens, pests, and weeds. Traditional methods for stress identification are labor-intensive and prone to inaccuracies. This paper presents a DEEP-CNN model designed for the multi-class classification of biotic stress in paddy crops. The proposed model utilizes preprocessing techniques and hyperparameter tuning to achieve optimal performance. Evaluations on the Paddy Doctor Dataset demonstrate the model's superior accuracy of 94.4%, outperforming existing state-of-the-art approaches. This research highlights the potential of deep learning for precision agriculture, providing an efficient solution for the timely detection and management of crop stressors.