Skip to content
SDGs Center UDINUS

Evidence record · UDINUS · SDGs

Hybrid SIFT-DCT Approach for Face Matching of Buddha Statues: Addressing Negative Similarity Metrics with S-Hashing

Rektorat UDINUS 2025 Published 2026-10-30

Evidence record · UDINUS · SDGs

Unit
Rektorat UDINUS
Period
2025
Goals
Record
578453d2-2ff9-4ba5-84fd-b2b315584bef
Published
2026-10-30

Universitas Dian Nuswantoro, through the activity participants Marlinda, Linda; Budiman, Fikri; Basuki, Ruri Suko; and Fanani, Ahmad Zainul, conducted research on image processing for the preservation and analysis of art and cultural heritage. The activity addressed the challenge of accurately detecting highly similar Buddha faces under variations in lighting, rotation, and minor facial differences. The researchers proposed a Content-Based Image Retrieval framework integrating Discrete Cosine Transform and Scale-Invariant Feature Transform to improve face-matching accuracy.

The system was tested on a database of Buddha images characterized by intricate textures and fine details. Within the method, DCT extracted global texture representations, while SIFT captured localized structural features. The results showed that DCT effectively encoded global texture characteristics. SIFT enhanced local feature detection, but it struggled to differentiate Buddha faces with extremely high similarity. The activity also identified instability in texture similarity computation. Chi-Square Similarity produced a -39.44% value for certain statues because of noise, artifacts, and lighting inconsistencies.

These findings indicate the importance of robust preprocessing techniques and refined similarity metrics for improving retrieval consistency. Overall, the hybrid DCT-SIFT approach improved the accuracy and robustness of Content-Based Image Retrieval systems in historical artifact datasets. Further research was identified as necessary to optimize preprocessing steps, integrate adaptive feature selection, and explore more stable similarity measurement techniques. In relation to Sustainable Development Goal 11, the activity contributes research concerning the analysis and preservation of cultural heritage through image-processing methods.

Reference https://www.scopus.com/pages/publications/105012774876?origin=resultslist