Batik Image Retrieval System Using Self Organizing Map
9 Pages Posted: 21 Sep 2018
Date Written: August 1, 2018
This paper describes an image retrieval system for batik motifs from Yogyakarta, Solo, and Lasem region using self organizing map method. The system receives user-uploaded batik image and returns 10 images of batik from Yogyakarta, Solo, and Lasem from the database with the closest distance that is computed using self organizing map. The experiments cluster the batik images into 13 clusters based on the 16 features extracted using gray level co-occurrence matrices (GLCM) to represent the texture of batik. In addition, color moment features are also considered as feature extractor to represent the color values of batik images. The average precision and recall for the experiments with both GLCM and color moment features are 0.5 and 0.082, respectively. Meanwhile, using only GLCM gives the result of 0.377 average precision and 0.060 average recall.
Suggested Citation: Suggested Citation