The production of agriculture and the security of food supply worldwide are seriously threatened by plant diseases. For mitigation strategies and timely interventions to be put into place, these diseases must be accurately and promptly diagnosed. In computer vision and image processing, deep learning methods have made promising strides in the last few years. This research uses deep learning techniques to investigate the automatic detection and categorization of plant diseases from leaf pictures. Applying transfer learning, we use pre-trained convolutional neural networks (CNNs) like ResNet50 and MobileNetV2, with over 54,000 images of various crop species and disease types from the PlantVillage dataset. Following rigorous testing and assessment, our method shows promise for agricultural applications with good accuracy, precision, recall, and F1 scores.