Batik Pattern Identification using GLCM and Artificial Neural Network Backpropagation

Abstract

Batik is a craft that has high artistic value and has become part of Indonesian culture. Preservation of batik must continue to be carried out so as not to be eroded by the times. One of the efforts to preserve batik is by digitizing batik patterns. With the diversity of batik patterns in Indonesia, digitizing motives is not only saved it in digital format but also how to retrieve and identify the batik patterns. In this study, we proposed a classification of Karawang batik pattern using backpropagation artificial neural networks and gray level co-occurrence matrix (GLCM) features. The step of the classification process begins with changing the image of batik from the color image to the grayscale image. Next, image segmentation is carried out to separate the batik pattern with its background using the Otsu method. Segmented images are extracted from GLCM features that are used as a feature in the classification process. Based on the results of testing using 50 test images, the accuracy value was 80%, the precision value was 91%, and the recall value was 83%.