This thesis deals with the application of convolutional neural networks (CNNs) for the classification of corn leaf diseases through digital image processing. The work is focused on the development of a deep learning model using CNN for the recognition and classification of three specific corn leaf diseases: gray leaf spot, rust, northern leaf blight, with the help of a publicly available image dataset. CNN was chosen for its ability to recognize patterns in images, which makes it ideal for plant disease classification. At the beginning, the theoretical framework of machine learning and deep learning is presented, including a more detailed treatment of the CNN architecture consisting of convolutional layers, pooling layers and fully connected layers, with reference to specific activation functions. The model development process begins with data collection, where images of corn with diseases are prepared and divided into training, validation, and testing sets. Data augmentation was applied to increase the number of samples, allowing the model to be trained on a larger number of images in various variants. The CNN architecture, implemented in the Python programming environment using Tensorflow, showed high accuracy during testing, reaching 97.96% accuracy on the test data. The obtained results demonstrated the model's high reliability in recognizing diseases, with minor deviations in specific cases due to the significant similarity between diseases. In conclusion, CNN proved to be very effective in the classification of corn leaf diseases and has the potential for further application in agriculture for other types of crops.