Intangible cultural heritage (ICH) is an important part of traditional culture, carrying deep historical and cultural values. With the advancement of science and technology, the application of Artificial Intelligence (AI) technology provides a brand new idea and method for the digital translation of non-heritage culture. This study explores the method of translating non-heritage ethnic graphic designs based on AI technology, and proposes an innovative path combining shape syntax and parametric design. Samples of Xiangxi Miao embroidery patterns are digitally collected, and image processing techniques such as bilateral filtering and rolling bootstrap filtering are used to preprocess the sample patterns so as to improve the accuracy of pattern extraction. The study uses the improved Canny operator for pattern extraction, and the experimental results show that the method in this paper has a significant advantage in the effective number of patterns extracted and the accuracy, and the extraction accuracy reaches 96.95%, which is far more than the traditional method. In order to verify the effectiveness of the method, the study also conducted expert scoring and user experience tests. The results show that the innovative design solution based on AI technology shows significant advantages in terms of satisfaction, visual experience and aesthetics of the work, and the comprehensive user score is 3.82 points higher than that of the traditional design solution, and the difference is statistically significant. The results of the study show that AI-assisted design not only enhances the artistic expression of traditional ethnic graphics, but also strengthens the user’s sense of identity and experience of traditional cultural works.
