Singular Value Decomposition in Machine Leaning for Image Compression in Vocational Tourism Batik Archiving

Batik archiving, Image compression, Machine learning, SVD, Vocational tourism

Authors

February 23, 2026
November 25, 2025

The digital archiving of batik products in vocational tourism environments requires efficient image compression techniques that maintain critical visual information, including complex motifs, color patterns, and texture details. This study aims to investigate the application of Singular Value Decomposition (SVD) as a machine learning based approach for image compression in the digital archiving of batik products from the Sundhullangit Batik Vocational Tourism Village. An experimental research design was adopted using digital batik images obtained through direct image acquisition. The research stages comprised image pre-processing, image compression using a truncated Singular Value Decomposition model with varying rank values, and reconstruction of the compressed images. The performance of the compression model was evaluated using objective image quality metrics, namely Mean Squared Error, Peak Signal-to-Noise Ratio, and Structural Similarity Index, while compression efficiency was measured using the compression ratio. The results indicate that higher rank values enhance reconstructed image quality, reflected by lower reconstruction error and higher structural similarity, but reduce compression efficiency. Conversely, lower rank values achieve higher compression ratios at the cost of reduced visual fidelity. Overall, the findings demonstrate that Singular Value Decomposition offers an effective balance between image quality preservation and data size reduction. This study concludes that the proposed method is suitable for supporting sustainable and high-quality digital archiving of batik products within vocational tourism-based cultural heritage systems.