MobileNetV2-based StrawDet Model for Early Detection of Strawberry Leaf Diseases

Strawberry leaf diseases are a foremost danger to the production and quality of strawberries grown worldwide. For disease management techniques to be effective, accurate and timely illness categorization is an essential step. Although useful, traditional illness detection techniques are frequently inefficient and not very scalable due to the high time complexity in early disease detection. Utilizing CNN models makes it possible to classify strawberry leaf diseases automatically and accurately using the visual symptoms that are collected in images. Research on the categorization of strawberry leaf diseases has the potential to transform agricultural disease control methods. For the classification of strawberry leaf diseases, a MobileNetV2-based StrawDet model has been developed and the observations indicate that epoch 15 exhibits the minimum loss, with a value of 0.0400 in the training phase and 0.0601 in the validation phase. Additionally, the training phase achieves a maximum accuracy of 99.21%, and 97.55% during the testing phase has been observed.