Plant diseases significantly affect agricultural productivity and pose a serious threat to food security. Among fruit crops, mango is highly susceptible to various leaf diseases that reduce both yield and quality. Early and accurate detection of such diseases is essential for effective crop management. In this study, a Convolutional Neural Network (CNN)-based approach is proposed for the automated classification of mango leaf diseases using images collected from real field conditions in Malda, West Bengal. The dataset consists of 980 images categorized into four classes: Healthy, Anthracnose, Red Rust, and Powdery Mildew. To improve model generalization, data augmentation techniques such as rotation and scaling were applied. The dataset was divided into training and testing sets in an 80:20 ratio. The proposed CNN model was trained for 40 epochs with a batch size of 35. Experimental results show that the model achieves a test accuracy of 90.36%, with precision, recall, and F1-score of 0.90 each, indicating balanced and reliable performance. The training and validation curves demonstrate stable learning behavior without significant overfitting. These findings confirm that the proposed approach is effective for accurate identification of mango leaf diseases under real-world conditions. The developed system has the potential to support farmers and agricultural experts in early disease diagnosis, thereby improving crop health and productivity. This work can be further extended by incorporating larger datasets and deploying the model in real-time applications.
Sulekha Kundu (Wed,) studied this question.