Digital Modeling and Cultural Inheritance of Chinese Opera Costume Patterns and Textiles Based on Image Recognition Algorithm

Qin Liang, Shidong Huang

Article
2026 / Volume 9 / Pages 19-41
Received 22 July 2025; Accepted 9 October 2025; Published 16 January 2026
https://doi.org/10.31881/TLR.2026.019

Abstract
Digital modeling of opera costume textiles plays an increasingly important role in garment preservation, textile engineering, and intelligent fabric design, where accurate pattern extraction, structural restoration, and weaving adaptability are critical challenges. Opera costume fabrics are characterized by dense ornamental patterns, multi-layer color arrangements, and strong coupling between visual appearance and textile structure, which makes conventional fabric modeling and garment pattern reconstruction inefficient and error-prone. To address these challenges, this study proposes an integrated digital textile modeling framework oriented toward costume fabric analysis and textile pattern engineering. A multi-color-space pattern recognition strategy is employed to enhance the extraction of fabric motifs and structural boundaries from high-resolution costume textile images. Based on the extracted fabric patterns, semantic associations between textile motifs, costume roles, and structural attributes are established to support textile classification and pattern interpretation in garment design. Furthermore, a structure-aware pattern modeling method is developed to ensure weaving feasibility, enabling the transformation of recognized patterns into textile-compatible representations suitable for jacquard weaving and virtual garment simulation. Experimental results demonstrate that the proposed method significantly improves the accuracy and structural integrity of costume fabric pattern segmentation while maintaining high consistency with traditional textile craftsmanship. The generated textile patterns show strong adaptability in digital fabric simulation, garment visualization, and weaving-oriented production workflows. This study provides a practical technical solution for digital garment pattern reconstruction, textile engineering applications, and intelligent preservation of trad.

Keywords
digital fabric modeling, image recognition, opera costumes, cultural knowledge graph, multi-scale segmentation