Decoding Lasem Batik: A Feature-Based Analysis of Motif Structure and Visual Composition

Lasem batik is the traditional, hand-drawn textile art form originating from Central Java, Indonesia. Renowned for its complex designs and cultural richness, Lasem batik has great heritage value. Not only for historical record but also for creativity in contemporary design, preserving and comprehending the motifs are essential. This paper intends to investigate computational method for examining the structural patterns present in Lasem batik motifs. Examining 5,879 motifs from five different Lasem batik variations, the aim is to find underlying design characteristics, categorize motif changes, and build a digital framework for preserving and maybe regenerating these patterns using Artificial Intelligence. A mix of machine learning and statistical methods was used to perform this analysis: Feature Extraction, Dimensionality Reduction, Clustering, Pattern Analysis, Evaluation and Visualization to display the motif relationships in two-dimensional space, therefore supporting the interpretation of the clustering outcomes. The computer study showed obvious patterns and structure in the Lasem batik motif dataset. Clusters formed based on shared visual attributes, with certain groups emphasizing symmetrical arrangements. The evaluation metrics confirmed that the clustering was meaningful and not arbitrary. This study demonstrates the effectiveness of combining machine learning techniques with traditional art analysis to uncover structural insights in heritage designs like Lasem batik. Beyond digital preservation, the findings pave the way for AI-assisted motif generation, opening possibilities for integrating traditional motifs into contemporary fashion, interior design, and digital art.