<![CDATA[This research discusses the application of the ResNet architecture, a deep learning algorithm based on Convolutional Neural Networks (CNN), for classifying rice grains through digital images. The study highlights the importance of automatic classification in quality control and the rice processing industry, as manual methods are subjective and inefficient for large data volumes. A dataset of rice grain images labeled at the individual grain level was created, and image preprocessing techniques such as normalization and augmentation were applied to improve training data quality. The ResNet model with several configurations was trained to recognize visual features such as shape, color, and texture of rice grains. Evaluation results indicate that the ResNet model achieves a reasonable accuracy of around 80%, with classification errors mainly occurring between visually similar rice varieties. The research suggests expanding the dataset and optimizing hyperparameters to further enhance model performance. The findings contribute to the development of AI-based systems for automated rice grain classification and quality inspection in the food industry.]]>
Copyrights © 2026
