Classification of Ulos Based on Type and Ethnic Origin Using the Support Vector Machine Method with GLCM Feature Extraction

Abstract

Ulos is a traditional woven fabric of the Batak tribe in Indonesia that has deep cultural and symbolic meanings usually used in various traditional ceremonies such as birth, marriage and death. This challenge is compounded by the lack of ulos introduction in the formal education curriculum and exposure to modern lifestyles. To address this issue, this research proposes the development of an accurate ulos classification system. This system combines Support Vector Machine (SVM) for classification and Gray-Level Co-occurrence Matrix (GLCM) for texture feature extraction. This approach is expected to effectively identify ulos types and the origin of Batak sub-tribes (Toba, Simalungun, Karo) based on visual and textural features, while utilizing technology to preserve and promote ulos culture. The results showed that the combination of GLCM and SVM successfully built a good ulos classification model, with the best model achieving an accuracy of 92.72% and F1-Score of 92.57%, indicating the model's excellent ability to detect ulos. The website system built for ulos classification uses local deployment. Suggestions for further development include the implementation of remote deployment for wider access and the addition of more diverse ulos patterns into the dataset.