Abstract With the development of virtual reality (VR) technology and augmented reality (AR), the communication media and expression of culture have been further expanded. In this paper, VR and AR technologies are applied to the digital IP character design of Min cultural heritage, focusing on the image generation and action recognition processes in the digital IP character design process. The LAFITE model is used to generate the digital IP image of Min cultural heritage, and after the pose representation of the digital IP character, the multi-class support vector mechanism is used to construct the action recognition model. The model tests proved that the FID and IS produced by the LAFITE model are superior to those produced by other traditional models by 30% to 316% and 6% to 48% respectively. The output images of the Min cultural heritage digital IP characters are also of better quality. The MSVM model exhibits a high recognition rate for various actions of the IP characters, with each index value exceeding 93%, thereby facilitating effective interaction and enrichment of digital IP characters. The image output and action recognition model proposed in the study can promote the innovative design of digital IP characters of Min culture and enhance the digital creative expression and interactive forms of Min culture.