Identification of Batik Motif Based Deep Learning-Convolutional Neural Network Approach

Abstract

Batik, a rich Indonesian cultural heritage, boasts a diverse array of motifs, each reflecting the unique philosophy of different regions. However, this diversity can make it challenging to distinguish between various batik patterns. This study aims to identify batik motifs using the Convolutional Neural Network (CNN) method. This research dataset comprises 521 digital batik images, encompassing five distinct motifs: Betawi, Cendrawasih, Kawung, Megamendung, and Parang. The data underwent a rigorous processing pipeline, including pre-processing, image segmentation, and feature extraction using Gray Level Co-occurrence Matrix (GLCM). Subsequently, a CNN model was employed for classification. The experimental results yielded an impressive average accuracy of 99.2% in identifying batik motifs. This outcome underscores the efficacy of deep learning, particularly CNNs, in recognizing and categorizing intricate batik patterns. This study may expect to serve a foundational step towards the development of advanced, automated batik recognition systems.