Classification of Batik Motifs Image Using Deep Learning Algorithms
Thebestindonesia.com – Batik is Indonesia’s cultural heritage which is caused not only by the batik itself but also by the manufacturing technique. Batik was originally an art form of Javanese royalty. However, there is an opinion that batik was famous among the nobility and ordinary people. Skills are needed to process it in batik because it requires a tool called canting and can only be done by women through the visual sense and touch of batik cloth, where historical and cultural meanings can be conveyed from generation to generation, passed on, and preserved. In addition, Batik contains social, cultural, and economic values that protect the dignity of the Indonesian nation, and batik is one of the national identities.
Batik has high artistic value and is an ancestral heritage that has become part of Indonesian (especially Javanese) culture. The word “batik” comes from the Javanese word “tik,” which means “dot/Matic” (verb, make a point), which later developed into batik. One of the ancient art forms of dyeing batik cloth is the block dye technique using wax. According to the Big Indonesian Dictionary, batik is a specially made painted cloth that is used or waxed and finished through a specific process. Batik is one of Indonesia’s handicrafts and cultural heritage, which has high artistic and economic value. Batik is a cloth-shaped craft with particular motifs, usually used as clothing material. Batik was also recognized by the United Nations Educational, Scientific, and Cultural Organization (UNESCO) in 2009. Batik has many unique motifs. Batik motifs are diverse, primarily written batik because batik experts make them. The following are the five most commonly used batik motifs: Ceplok, Kawung, Parang, Megamendung, and Sidomukti.
Batik has a profound meaning and various motifs. Almost every region in Indonesia has unique motifs, and many artists develop new motifs to keep batik growing and sustainable from generation to generation. However, many modern Indonesians need to learn the types and classifications of batik they use. Some even use batik because their ancestors classified it. In addition, most of the knowledge about it is only sometimes passed on to their children and grandchildren. Thus, the batik image needs to be classified because it has different motifs and can be taken from various sources. Batik motifs continue to develop from time to time so that knowledge about it is restarted.
Read also : Image Classification Using Transfer Learning and CNN
The challenge of classification does come from not only the variety of batik motifs but also the problem with the image of batik, which has various conditions because it can come from the internet or self-snapshots as well as the background to the need for technology that can classify batik. For example, batik images can be found on the internet in various scales and resolutions. This happens because it has no control over the uploaded batik images. One of the technologies that can be used is the application of Artificial Intelligence, especially in deep learning algorithms such as Convolutional Neural Networks (CNN). Deep learning can solve the core problem of learning representations by introducing representations expressed in other simple representational forms. Deep learning enables computers to generate complex concepts from simpler ones. Deep learning is a branch of machine learning that uses artificial neural networks to solve problems on large data sets. Deep learning techniques provide a robust architecture for supervised learning by adding a layer. The training model can represent well-labeled imagery.
CNN is one of the deep learning algorithms representing the development of multi-layer perceptrons (MLP). CNN can detect and recognize objects in an image. It can also outperform traditional computer vision and pattern recognition methods, such as object detection, classification, image segmentation, and text recognition. CNN can classify labeled data that represents its neurons in a two-dimensional format. Therefore, CNN is one of the most popular and accurate models implemented in image classification. Several studies have classified batik motifs using CNN and produced a reasonably high accuracy. So, based on the research that has been done, it shows that the CNN model can be used for batik classification. Still, the CNN model made you have yet to obtain reliable accuracy in batik classification.