Plant Disease Detection using Deep Learning Models in Precision Agriculture

Plant diseases affect agriculture worldwide due to their lower crop yields and increased pesticide need. This study explores how deep learning is used for plant disease diagnosis, emphasizing its role in precision agriculture. Through an eye on classification accuracy and processing efficiency, the study compares models like ResNet50, EfficientNetB0, InceptionV3, DenseNet121, and MobileNetV2. The most successful model is ResNet50, which has an accuracy of 98.85%—using a dataset of 1,700–7,400 photos of high-demand crop's like cotton, tomato, eggplant, and more. DenseNet121 and EfficientNetB0 provide comparable performance with slightly different computational trade-offs. Model performance, dataset characteristics, and the impact of architectural decisions on detection accuracy are all included in the study. It also highlights practical applications, tackling real-time deployment constraints and dataset limitations. The study emphasizes how deep learning models can enhance sustainable farming methods, reduce pesticide reliance, and improve early illness diagnosis. This study helps practitioners and researchers construct efficient plant disease detection systems by offering insightful information on model selection and optimization