Research on Generative Design of Chinese “Dunhuang Zaojing” Patterns Based on AIGC

Abstract

With the rapid development and increasing maturity of AI technology, AIGC is being applied and continuously developing in an increasing number of fields. Secondly, the inheritance and redesign of traditional Chinese patterns, particularly “Dunhuang Zaojing” patterns, in the new era has become a new direction for the renewal of traditional patterns. Currently, the use of AIGC technology to modernize traditional patterns has become a new approach to achieving design diversification. Therefore, this study aims to explore the inheritance and innovative design paths of“Dunhuang Zaojing" patterns based on AlGC (Midjourney). First, a “Dunhuang Zaojing” pattern data set was constructed based on literature, four generation methods were designed, and an evaluation system consisting of four dimensions and eight indicators was constructed. Second, "Dunhuang Zaojing" patterns were generated using Midjourney using the four methods, while the generation time was recorded. Finally, the generation time and evaluation results were analyzed. The generated results achieve innovations in detail and composition, offering new avenues for the digital display and modern design transformation of traditional "Dunhuang Zaojing” patterns. They lower the creative barrier while preserving the core structure and contemporary color characteristics. They also clarify the advantages and optimization directions of AlGC in traditional pattern generation, providing a reference experimental framework and evaluation approach for subsequent intelligent design research on similar intangible cultural heritage elements.