Palm Tree Leaf Disease classification using Hybrid Deep Learning model

Background Palm trees are one of the main components of global ecosystems and economies, but leaf spot diseases in palm trees can significantly harm their health and productivity. The early discovery of these diseases is a vital step that is essential for effective disease management and prevention. This paper combines Convolutional Neural Networks (CNNs) for feature extraction and an XGBoost Classifier to propose a new palm tree leaf disease classification. Our proposed method makes use of CNNs, which are suitable for extracting the features while at the same time extracting the most discriminative information from palm tree leaf images. It uses XGBoost to classify regular and infectious (spotted) leaves by the features. Our method was validated using a large dataset of images, achieving an accuracy of 0.86, proving our approach's effectiveness and robustness for palm tree disease detection. We outperform traditional methods and standalone models, demonstrating the promise of our approach for practical palm tree disease management and agricultural uses