Effective detection of plant leaf diseases is crucial for enhancing agricultural productivity and minimizing crop losses. This study examines the integration of Convolutional Neural Networks (CNNs) with morphological operations, specifically skeleton and opening methods, for the precise identification of plant leaf diseases. CNNs are used due to their high capabilities in feature extraction and image classification. The CNN model trained on a dataset consisting of healthy and diseased leaf images exhibits robust performance in disease detection. Additionally, morphological operations are applied to facilitate a detailed analysis of leaf vein patterns and affected areas, enabling a more in-depth examination of the leaf structure. A comparison of the results obtained from CNN models using skeleton and opening operations reveals that the opening operation achieves a significant 80% success rate in disease detection. This finding demonstrates that the effectiveness of the opening method in detecting plant diseases surpasses that of the skeleton method. As a result, this integrated approach has the potential to increase agricultural productivity by providing an advanced and reliable solution for plant health monitoring.