Comparison Of Classification Method On Lombok Songket Woven Fabric Based On Histogram Feature
Rina Aprianti, Kristhina Evandari, Ricardus Anggi Pramunendar, Moch Arief Soeleman · 2021
Songket fabric has repeated geometric shapes in one type of fabric; some parts of the forming motif. Can distinguish motifs based on the shape of the motif, the density of the motif, the number of colors, and the position of the motif on the songket cloth. One area that is famous for its songket woven fabric is Lombok. Lombok people's awareness of the songket motif is still minimal, and the difference between one motif and another is still unknown. Previous research only used one algorithm. This study uses a method by applying four classification algorithms, where the image is enlarged and a comparison is made of each pixel size used. The results obtained from each image size produce the other highest accuracy values, where the highest product in the Naive Bayes algorithm is 90% with a pixel size of$100\times 100$. The size is$300\times 300$pixels, the Naive Bayes algorithm for the highest accuracy value is 80%, while for the$400\times 400$size the highest value is the Decision Tree algorithm at 90%. Where from this comparison, the algorithm with the highest accuracy value is Naive Bayes. The results given are expected to help and facilitate the community in recognizing songket motifs.