<![CDATA[The circulation of beef and pork products that are difficult to distinguish visually can create challenges for consumers, making an automated meat identification system necessary. This study aims to develop an image classification model for beef and pork using the Convolutional Neural Network (CNN) method with the DenseNet-121 architecture and to implement it in a Streamlit-based web application. The dataset used in this study consists of 6,000 images, comprising 3,000 beef images and 3,000 pork images collected from two different dataset sources. The dataset underwent several preprocessing stages, including resizing, contrast enhancement, normalization, and data augmentation, and was subsequently divided into training, validation, and testing sets with a ratio of 70:15:15. The results show that the DenseNet-121 model is capable of classifying beef and pork images with excellent performance. Based on the evaluation using a confusion matrix and classification report, the model achieved an accuracy of 97.89%, with high precision, recall, and F1-score values for both classes. The trained model was then deployed in a web application that allows users to perform classification through image uploads or direct image capture using a camera. Based on these findings, it can be concluded that the DenseNet-121 architecture is capable of classifying beef and pork images with high accuracy and has the potential to be utilized as a practical tool for meat type identification.]]>
Copyrights © 2026
