Intelligent Reconstruction of Visual Knowledge for Intangible Cultural Heritage Patterns
Traditional costume patterns carry intact visual tacit knowledge of folk culture. Digital collection generates damaged patterns and breaks complete knowledge chains. Existing text-based heritage archives cannot store spatial and color knowledge of patterns. Conventional image restoration only optimizes pixel effects and ignores standardized knowledge preservation. This work constructs a unified dual-branch generative framework for pattern visual knowledge completion. The structural branch adopts graph relation constraints to repair topological knowledge of repeating units and symmetry axes. The color branch uses global color wheel coding to realize cross-scene semantic knowledge adaptation. A two-way feedback mechanism realizes mutual calibration of structural and color knowledge. Tests on 312 ethnic pattern samples show the model improves topological knowledge integrity and color knowledge stability. The framework outputs standardized labeled knowledge assets for heritage archive management. This research expands the processing paradigm of visual tacit knowledge.
Authors
- Zhiwei XIAO (ORCID: https://orcid.org/0009-0008-1515-1094)
Publication Details
- Journal
- International Journal of Knowledge Management
- Published
- 2026-11-17
- DOI
- https://doi.org/10.4018/ijkm.427936
- Primary Topic
- Image Processing and 3D Reconstruction
- Type
- article
- Field-Weighted Citation Impact
- 0.00