<![CDATA[Batik is an Indonesian cultural heritage with diverse patterns and deep philosophical meanings. The advancement of artificial intelligence enables automatic recognition of batik motifs, supporting cultural preservation. This study develops a batik motif classification system using the Convolutional Neural Network (CNN) architecture, specifically ResNet18. The dataset consists of 1,427 images from 14 types of batik motifs, including Parang, Priangan, Pring-Sedawung, Kawung, and Megamendung. Preprocessing steps involved resizing images to 128x128 pixels and splitting them into training and testing sets across five scenarios. Experimental results indicate that the fourth scenario (60:40) achieved the best performance with 80% accuracy. These findings demonstrate that ResNet18 is effective for batik motif classification, although further improvements may be achieved with larger datasets and advanced augmentation techniques.]]>
Copyrights © 2025
