Mango Leaf Disease Detection Using Deep Feature Extraction and Machine Learning Methods: A Comparative Survey

Abstract

Plant diseases pose a significant threat to the quality and quantity of agricultural production, with leaf diseases being particularly detrimental to plant growth and yield. In the near future, ensuring access to affordable and safe food will become one of the most pressing global challenges. As a result, the early detection of plant diseases is crucial for both economic stability and food security. Detecting and monitoring diseases in mango leaves, however, is a complex task when relying solely on visual inspection. This study seeks to address this challenge by utilizing image processing and deep learning techniques to detect mango leaf diseases. We extracted deep features from mango leaf images using several prominent architectures, including Darknet19, Xception, SqueezeNet, MobileNetv2, DenseNet201, GoogLeNet, ResNet18, VGG16, and AlexNet. These features were then classified using machine learning algorithms such as decision tree, linear discriminant analysis, naive Bayes, support vector machine, k-nearest neighbors, and ensemble classifiers. Our findings demonstrate an improvement over existing results in the literature, with detailed experimental results presented within the article. © 2025, TUBITAK. All rights reserved.