A CONV-EGBDNN Model for the Classification and Detection of Mango Diseases on Diseased Mango Images utilizing Transfer Learning

Abstract

Mango fruits are highly valued for their taste, flavor, and nutritional value, making them a popular choice among consumers. However, mango fruits are susceptible to various diseases that can significantly affect their yield and quality. Therefore, accurate and timely detection of these diseases is crucial for effective disease management and minimizing losses in mango production. Computer-aided diagnosis techniques have emerged as a promising tool for disease detection and classification in mango fruits. This study adopts an image classification approach to identify various diseases in mangos and distinguish them from healthy specimens. The pre-processing phase involves a Wiener filter for noise removal, followed by Otsu's threshold-based segmentation as a crucial operation. Subsequently, features are extracted by implementing the ResNet50 model. The proposed model was experimentally verified and validated, demonstrating optimal results with an accuracy of 98.25%. This high accuracy rate highlights the effectiveness of the XG-Boost classifier in accurately categorizing mango images into different disease categories. The experimental results strongly support the potential practical application of the model in the agricultural industry for disease detection in mango crops.