<![CDATA[Batik Solo has a rich variety of motifs with high philosophical value, but the process of identifying motifs is still largely done manually, making it subjective and prone to error. This study aims to develop an automatic classification system to distinguish four Batik Solo motifs, namely Parang, Kawung, Truntum, and Sekar Jagad, using the Convolutional Neural Network (CNN) method based on the VGG16 architecture with a transfer learning approach. The dataset used consists of 280 batik images divided evenly into four classes (70 images per class), where data limitations are overcome using ImageNet pre-trained weights, freezing all convolution layers, and applying data augmentation to reduce the risk of overfitting. The selection of VGG16 was based on the consideration that this study focused on evaluating feature extraction capabilities and analyzing the classification performance of Batik Solo visual patterns in depth, so VGG16 was used as a stable and interpretative baseline model, not for the purposes of computational efficiency or mobile implementation. The training process was carried out for 50 epochs with a data division of 60% training data, 20% validation data, and 20% test data, and the test results showed an accuracy of 85.71% with average precision, recall, and F1-score values of 0.88; 0.86; and 0.86, respectively, where the Sekar Jagad motif performed the best, while the Truntum motif was the most challenging class due to its smooth and repetitive texture characteristics.]]>
