COMPARISON OF K-NEAREST NEIGHBOR AND SUPPORT VECTOR MACHINE CLASSIFICATION ALGORITHMS IN PATTERN RECOGNITION OF TAPIS FABRIC MOTIFS USING NON-GRAYSCALE LBP FEATURE EXTRACTION

Tapis fabric is a traditional garment of the Lampung people, made from cotton threads and adorned with silver or gold thread motifs. Tapis fabric is an important cultural heritage for the people of Lampung, Indonesia, with its motifs holding deep historical and symbolic meanings. The aim of this research is to develop a classification model for Tapis fabric patterns using K-Nearest Neighbors (KNN) and Support Vector Machine (SVM) algorithms. This involves utilizing Local Binary Pattern (LBP) without converting the images to grayscale, thereby preserving the color in Tapis fabric motifs. The goal is to compare the performance of the two algorithms based on accuracy, precision, recall, and F1-score metrics. The application of digital image processing technology, particularly through the use of LBP feature extraction and appropriate classification algorithms, provides a significant contribution to facilitating the identification and classification of Tapis fabric. This research focuses on the development and identification of classification techniques to more accurately and efficiently distinguish the complex and varied Tapis fabric motifs. In this study, the KNN algorithm was applied with various k values, while the SVM algorithm was tested with different kernels, including RBF, linear, polynomial, and sigmoid. The results indicate that the KNN algorithm with k = 3 achieved the best results with an accuracy of 94%, while the SVM algorithm with the RBF kernel achieved the highest accuracy of 84%. These results show that KNN is more effective than SVM in the context of Tapis fabric motif classification for this study.