Convolutional Neural Network untuk Klasifikasi Citra Makanan Khas Indonesia

Muhammad Dandi Darojat (Faculty of Computer Science, Brawijaya University)
Yuita Arum Sari (Faculty of Computer Science, Brawijaya University)
Randy Cahya Wihandika (Faculty of Computer Science, Brawijaya University)

Jurnal Pengembangan Teknologi Informasi dan Ilmu Komputer
Vol 5 No 11 (2021): November 2021

There are various types of special food in Indonesia that are still difficult to identify domestically and internationally. If many people know the special Indonesian food, then Indonesia will be increasingly recognized as well. In line with this, state revenue may increase as well. But in some cases, special food in Indonesia is still challenging to identify, especially for foreign tourists. This paper proposes an image classification system for Indonesian special food images using the Convolutional Neural Network algorithm (CNN) which is supported by several other methods and algorithms. Based on the experiments conducted eight times on 26 models, the best model was obtained with a test accuracy value of 0.6 and an evaluation accuracy of 0.91. This shows that the CNN is relatively good to be applied to the classification of special Indonesian food images.