Artificial Intelligence Untuk Identifikasi Motif Tenun Tradisional Sulawesi Tengah

Pusadan, Mohammad Yazdi
Laila, Rahma
Pratama, Septiano Anggun

Bomba: Jurnal Pembangunan Daerah
Vol 5 No 1 (2025)

Traditional weaving from Central Sulawesi, such as the motifs of Magau, Banua Oge/Souraja, and Tadulako, reflects deep cultural and historical values. However, the complexity of the patterns and motifs often makes manual identification challenging. This research employs an Artificial Intelligence (AI) approach using Convolutional Neural Networks (CNN) to automate the identification of these motifs. The AI model is trained using a diverse dataset of woven motif images and shows significant accuracy in classifying Magau, Banua Oge/Souraja, and Tadulako motifs. This research opens up cultural preservation and innovation opportunities in woven products with modern technology. The achieved result is the evaluation of the AI model using the following metrics: accuracy, precision, recall, and the confusion matrix. The accuracy obtained for each motif reaches 90%.