<![CDATA[Indonesia has a rich culinary diversity, encompassing various types of food from different regions. In the current era of technological advancement, the application of artificial intelligence has grown significantly across multiple sectors, including in the identification of Indonesian food images. This research provides the impact of various image preprocessing techniques on the AI-based food identification system. The preprocessing methods examined include Zero Component Analysis (ZCA), Histogram Equalization (HE), Contrast Stretching, and Image Sharpening. The evaluation of these preprocessing methods was conducted to determine which technique provides the best performance in assisting the identification of Indonesian food using a Convolutional Neural Network (CNN) with a ResNet-50 transfer learning model. Performance measurement was carried out using a confusion matrix by calculating Accuracy, precision, recall, and F1-score. The results of this research show that the use of the Image Sharpening method yields higher accuracy and precision on the testing data compared to other methods, those are 0.9748 and 0.98, respectively. Next, a high level of accuracy was also demonstrated by the Contrast Stretching method, with an accuracy score of 0.9712.]]>
Copyrights © 2025
