In order to realize the collision and fusion of Yi clothing culture and modern clothing design in Liangshan, China, this paper proposes the innovative fusion strategies of pattern cutting and color feature extraction for Yi clothing, respectively, and constructs a corresponding implementation model on the basis of artificial intelligence technology. Based on the UNet model, the pattern segmentation model is constructed by integrating multi-scale feature extraction and attention, embedding ASPP and ECA attention, and combining the Lovász-hinge loss function. The color network relationship model is constructed and optimized, using K-means clustering method to extract the image application color and other related color data, constructing and optimizing the color network relationship model. Using the Yi dress pattern cutting and color feature extraction technology, a series of clothes with the theme of “Yi Seeing and Loving” is designed by integrating the Yi dress cultural elements. Respondents think that the overall color and pattern matching style of the series of clothing designed in this paper is more inclined to gorgeous, traditional, coordinated, chic, modern and easy to match, and the evaluation value of the clothing style design and decorative detail design focuses on the interval of [3,5]. At the same time, the proportion of respondents who evaluated the degree of ethnic style in the [3,5] interval and the design effect in the [2,3] interval was as high as 80%.