Revolutionizing Cucumber Agriculture: AI for Precision Classification of Leaf Diseases

Cucumbers are crucial agricultural commodities worldwide, necessitating production enhancements and quality maintenance. However, several diseases can easily hamper cucumber production if not classified and detected early. This study proposes a modified deep learning approach for identifying and classifying cucumber diseases at early stages using the Kaggle dataset. We applied data augmentation techniques to enhance the robustness of our model. We replaced the original MobileNetV2 classifier with custom classifiers, including linear layers, batch normalization, and ReLU activations to identify the effective features. Our proposed model achieved a high accuracy of 96.81%, which surpasses the accuracy of recent existing models in cucumber leaf disease classification. In the future, we plan to use a larger dataset, and this model can be adapted to work with various IoT devices for deployment in real-time applications.