Deep Learning-Based Detection of Mango Leaf Diseases Using CNN Architecture

Mango leaf diseases pose a significant threat to mango production, impacting yield and quality. Accurate and early detection of these diseases is critical for effective management and sustainable agriculture. This research presents a deep learning-based approach for automatic mango leaf disease detection using convolutional neural network (CNN) architectures, including VGG16, VGG19, ResNet50, DenseNet121, and AlexNet. A comprehensive dataset of mango leaf images, representing various diseases such as Anthracnose, Bacterial Canker, and Powdery Mildew, was preprocessed and categorized. Each model was trained and evaluated using image data augmented for improved generalization, with class labels converted to categorical formats. Results indicate that DenseNet121 achieved the highest accuracy, with consistent precision and recall across disease classes, suggesting it as an optimal model for this application. This study highlights the potential of deep learning for precise and scalable plant disease detection, offering a promising tool for automated agricultural disease management. Future work will explore model deployment in mobile applications to support real-time disease detection in the field.