Plant diseases can cause widespread damage quickly if not detected early. Identifying infections promptly allows farmers to apply treatments more precisely, reducing chemical usage and protecting the environment. Tomatoes, one of the most common crops worldwide, are prone to leaf diseases that affect their productivity and quality. Manual detection methods are time-consuming, require significant labor, and depend on expert skills. This study presents an automated classification system for the early identification of tomato leaf diseases through image processing and deep learning, utilizing techniques such as preprocessing, dimensionality reduction, feature extraction, and classification. A VGG16 network has been applied to capture texture features, which are further processed by a classification model. This texture feature includes two main parts: an enhanced base term and a texture detail component, providing a detailed textural assessment. The model, trained on tomato leaf images from the Plant Village dataset, achieved an exceptional accuracy of 99.27%. The comparative analysis with existing methods demonstrates that the proposed approach outperforms traditional models in accuracy and efficiency. The findings highlight the potential of deep learning in precision agriculture, paving the way for real-time automated disease detection systems to support farmers and improve crop management.