<![CDATA[Diseases in maize plants are one of the main factors contributing to yield reduction, ultimately impacting food security and the sustainability of the agricultural sector. Technological advancements have enabled digital image processing to become a widely applied method in various fields, including object identification, pattern detection, and plant disease classification with the support of Artificial Intelligence (AI). This study utilizes maize leaf images as a dataset, which is then processed using the Convolutional Neural Network (CNN) method based on the ResNet-50 architecture. This architecture is known for its superior ability to extract deep visual features, thereby enhancing classification accuracy. The CNN model operates by identifying and analyzing key features in images, such as color, texture, and damage patterns on maize leaves, to detect the type of disease affecting the plants. The methodology in this research involves several crucial stages, starting from collecting maize leaf image datasets directly from agricultural fields, preprocessing data to improve image quality for optimal model training, and finally, training and evaluating the CNN model’s performance. The initial dataset consisted of 1,199 images, which, after image analysis, was reduced to 870 for processing. These were then split into 552 images for training, 87 for testing, and 261 for validation. The CNN model evaluation with 87 test images showed that the highest precision was achieved in the Rust Disease class (100%). Meanwhile, the highest recall was found in the Normal Leaf class (99%), indicating that the model correctly identified all normal leaf samples. However, the recall for the Fall Armyworm Disease class was 89%. This model accurately identifies damage patterns, making it a potential tool for automatic early disease detection in plants. Thus, this research can serve as a foundation for further developments in AI-based applications to enhance efficiency and accuracy in plant disease diagnosis, ultimately supporting agricultural productivity and sustainable food security.]]>
