Detection and Classification of Fruit Tree Leaf Disease Using Deep Learning

Plant disease identification is extremely important in agriculture since it is critical for boosting crop output. Visual plant disease analysis is a modern technique to handle this problem, following recent developments in imaging. In this study, we look at the challenge of plant disease detection which is visually done for identification of plant disease. Plant disease images, in comparison with other types of photographic images, are likely to have randomly dispersed lesions, varied symptoms, and complex backgrounds, making discriminative information difficult to capture. To facilitate plant disease recognition research, we had taken the Plant Village dataset with 13,347 images with 14 classes. Models were trained using the Plant Village dataset. The performance of EfficientNet architecture for classifying the plant leaf disease was compared against ResNet-50, Inception V3, AlexNet, and Xception deep learning algorithms in this analysis. The outcomes of the test dataset revealed that B3 models of the EfficientNet architecture had the greatest accuracy of 99.90 percent when related to other deep learning algorithm in the dataset.