Research on the Digitalization and Application Path of Ethnic Art Image Archives in Textile Product Design

Miao Li

Article
2026 / Volume 9 / Pages 255-270
Received 20 October 2025; Accepted 20 November 2025; Published 30 January 2026
https://doi.org/10.31881/TLR.2026.255

Abstract
This research establishes a systematic framework for textile design innovation by integrating digital archival methods with contemporary textile manufacturing technologies. It specifically investigates the application of digitized ethnic art archives for creating novel textile surface patterns, aiming to bridge cultural heritage with modern fabric design and production. Focusing on Miao ethnic art, the study details a methodology comprising three stages: (1) high-fidelity digitization and semantic analysis of traditional textile patterns to build a structured digital archive; (2) an AI-driven creative process utilizing StyleGAN2-ADA with transfer learning to synthesize novel, printable textile patterns from a limited archival dataset; and (3) an empirical evaluation of the resulting digitally printed textiles through market analysis and consumer surveys. A quantitative analysis of 500 traditional patterns informed the GAN model, which successfully generated 100 new textile designs. Validation through nearest neighbor analysis confirmed that the model produces stylistically consistent yet visually distinct variations without directly replicating training samples. Furthermore, a color fidelity test on digitally printed samples yielded a ΔEcmc of 2.4, verifying the physical reproducibility of the generated designs. A comparative consumer survey (n = 600) confirmed a significantly higher purchase intention (p < 0.001) for the expert-curated selection of textiles featuring these culturally grounded designs compared to generic prints. The discussion addresses the critical balance between design innovation and cultural appropriation and the role of digital technology in revitalizing traditional motifs for the modern textile industry. The findings validate a replicable pathway for transforming cultural assets into market-viable textile design concepts.

Keywords
digital textile printing, textile pattern design, computer-aided design (CAD), ethnic patterns, Miao art