Mango Leaf Detection Using Transfer Learning

Abstract

Mango cultivation plays a pivotal role in the agricultural economy of many countries, especially in tropical regions. However, the productivity and quality of mangoes are significantly affected by various leaf diseases, which often go undetected until they cause substantial damage. Traditional methods of disease identification are time-consuming and require expert knowledge, making them less feasible for timely intervention. The integration of deep learning techniques, particularly transfer learning, offers a promising solution for the early and accurate detection of mango leaf diseases. This project leverages the capabilities of transfer learning by utilizing pre-trained convolutional neural networks, specifically MobileNetV2, to identify and classify mango leaf diseases. By employing image preprocessing techniques and data augmentation, the model is trained to recognize patterns associated with different diseases. The system provides users with an intuitive interface to upload leaf images and receive instant diagnostic feedback, thereby facilitating prompt and informed decisionmaking in disease management. Keywords: Mango Leaf Disease, Transfer Learning, MobileNetV2, Deep Learning, Image Classification.