<![CDATA[Indonesia is rich in natural resources, including spices that are widely used as flavor enhancers in food and as traditional herbal medicine. This study aims to develop a web-based spice classification system using a Convolutional Neural Network with the DenseNet121 architecture. The dataset is multi-class and consists of 3,600 images representing 30 types of spices, including fennel, andaliman, tamarind, onion, shallot, garlic, coriander seed, star anise, clove, kaffir lime leaf, basil leaf, coriander leaf, bay leaf, ginger, cumin, cardamom, cinnamon, sappan wood, candlenut, cubeb, aromatic ginger, kluwek, turmeric, pepper, galangal, nutmeg, saffron, lemongrass, vanilla and sesame. The system was developed using the Cross-Industry Standard Process for Data Mining (CRISP-DM) which consists of six stages namely business understanding, data understanding, data preparation, modeling, evaluation and deployment. The dataset was divided into training and testing sets with four proportions 90:10, 80:20, 70:30 and 60:40 to compare model performance. The experimental results show that the 80:20 proportion achieved the best performance with 99% training accuracy, 95% validation accuracy and 95% accuracy as the main performance metric, and this model successfully classified all spice categories. The trained model was then deployed into a website using the flask framework which enables practical use for spice image classification. ]]>
Copyrights © 2026
