Along with the flow of clothing designer personnel, clothing style design technology will be lost, so the intelligent design of the clothing industry, with the development of artificial intelligence technology, has received great attention. In this paper, based on the real-time style migration algorithm to study and design the local pattern of clothing, based on the principle of real-time style migration of images with perceptual loss function, it is proposed to optimize the real-time style migration network by using the group normalization method to obtain the style migration image that is more in line with the visual habit. The fusion of the style migration image with the garment base map through Poisson fusion achieves the end-to-end intelligent design of garment local patterns. The AdaIN-based style migration network model designed in this paper has an average accuracy of 75.758% when trained. Then, the model is applied to the clothing fashion design to evaluate the effect of the garment. The average scores of the overall, front, side, and back evaluation of the dressing effect of the garment are 4.04, 4.15, 4.01, and 3.92, respectively. The garments show a better effect. AI technology will bring more innovative possibilities to the clothing fashion industry and inject new vitality into the fashion industry.