Deep Learning for Date Palm Disease Detection

Date palm trees are of great artistic, profitable, and environmental significance for numerous countries. In the Kingdom of Saudi Arabia, where the date palm is the national tree, dates are a key export and an essential element of local cuisine. Thus, conditions affecting date palm trees are veritably serious peril that significantly reduces their productivity and fruit quality. Traditional methods for detecting these conditions rely on mortal observation, which is frequently time- consuming, expensive, and prone to inaccuracies. This study employs convolutional neural network (CNN) pre-trained models to develop an automated and accurate disease detection system for date palm trees. A dataset of 5,094 palm leaf images, categorized into eight common diseases and healthy leaves, was collected from 10 farms in Madinah, Saudi Arabia. VGG16 and MobileNet were chosen for their balance between accuracy and computational effectiveness, making them suitable for real-world operations. The results demonstrate high accuracy, outperforming traditional methods. This exploration highlights the role of AI in advancing smart agricultural practices and contributing to sustainable date palm farming.