Motif-Sikka: A Curated Image Dataset of Traditional Sikka Ikat Weaving Motifs for Textile Pattern Analysis

This dataset was created under the hypothesis that image-based computational methods can effectively be used to classify and recognize traditional textile motifs, particularly Sikka ikat weaving patterns. Variations in lighting, device resolution, and motif structure are significant challenges in textile image analysis. By capturing these variations in structured dataset, we hypothesize that it will support the development of more robust and generalizable computer vision models for motif recognition and cultural heritage preservation.
This dataset is a subset of the Sikka-MoDiVers collection, consisting of 432 images from 24 verified classes of traditional Sikka ikat weaving motifs. Each class is represented by images captured under six natural lighting conditions:
Indoor morning, Indoor noon, Indoor night, Outdoor morning, Outdoor noon, Outdoor night
The images have been resized to 512×512 pixels to standardize them for machine learning tasks such as image enhancement, segmentation, feature extraction, and motif classification. The naming convention of image files reflects both the motif class and the lighting condition.
Each motif class is stored in its own folder, with clear labeling. The dataset is aligned with the motif classification from the Geographical Indication Certificate ID G 0000000564, issued by the Directorate General of Intellectual Property, Republic of Indonesia.
Included in this release:
1. A metadata CSV file (classes.csv) listing motif name, number of images, and lighting conditions.
2. A README file (README_v3.md) that describes the directory structure, naming format, and usage instructions.tudies involving traditional Indonesian textiles.
Potential Uses:
1. Deep learning model training for traditional textile recognition.
2. Cultural heritage digital archiving and motif classification.
3. Evaluation of enhancement techniques under varying lighting conditions.

How to Use:
Extract the dataset and refer to the metadata file for understanding the structure. Each folder contains images from a single class, suitable for direct use in supervised learning models.