When illustration meets traditional Chinese folk art, how to create a new vision with both cultural roots and contemporary aesthetics is the goal of the research. In this study, we design a set of combined paths for target recognition and style migration. On the one hand, the key features of line curvature, texture pattern (LBP) and spatial frequency are extracted from the images. On the other hand, the style migration model is innovatively adapted to address the asymmetry in the amount of information between folk art and real photographs. Two asymmetric generators with different capabilities are designed and equipped with feature-level cyclic consistency loss and saliency edge loss. The average accuracy (mAP) of the target recognition model in this paper is 57.82% and 53.13% on the mural and illustration datasets, respectively, which outperforms the mainstream detection models. The asymmetric style migration model generates images with a structural similarity of SSIM = 0.7087 to the original photographs, and a content classification accuracy of 70.72%, both of which are substantially better than the comparison methods. Its social acceptance was further explored through a questionnaire survey, with more than 93% of respondents expressing acceptance or great acceptance of this form of integration. They are most looking forward to interacting with it in the form of offline exhibitions (72.08% support) and cultural products (61.04%).
