Bell Pepper Plant Disease Detection using Deep Learning Models

Early identification of diseases in bell pepper plants at the appropriate time creates two benefits: improving crop output and reducing economic losses for agriculture. This research develops a deep learning system utilizing ResNet50, MobileNetV2, and InceptionV3 as core Convolutional Neural Network (CNN) architectures to examine bell pepper leaf health conditions. The researchers collected images from both healthy and diseased leaves for training and evaluating the processing model. Precision, recall, accuracy, and F1-score served as the assessment metrics. Each of the three CNN models presented exceptional results from their classification tasks. The detailed feature tracking capacity of ResNet50 was available, but MobileNetV2 achieved fast operation with efficient performance, and InceptionV3 demonstrated its reliability in multi-scale feature learning.