Machine Learning-Based Image Classification of Lampung Tapis Fabric Patterns Using CNN

Lampung Tapis is a traditional fabric that has high cultural value with various patterns and motifs. Modern tapis fabric patterns are currently continuing to develop by modifying several parts of the traditional patterns so that the number of pattern variations is increasing. This research aims to test a convolutional neural network (CNN)-based machine learning model that can identify Lampung tapis fabric patterns from digital images. Image classification with CNN is used to classify tapis fabric images into certain categories based on texture, shape, and color features. Experimental results indicate that the CNN model can classify tapis fabric patterns with good accuracy, reaching more than 88%, although there are challenges in some classes with high pattern variation. There are several opportunities to conduct further research; the performance results produced are still <tex xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">$&lt;90 \%$</tex>. This indicates improving performance optimally in terms of image classification, especially related to Lampung tapis fabric motifs. and try to use other CNN models such as VGG16, VGG19, ResNet50, or InceptionV3 or by using the deep learning model.