In this paper, the SIFT algorithm is used to detect as well as match the feature points of the artwork, select the centroid of the artwork as well as the support domain, and produce locally approximated surfaces in the support domain to perform a 1:1 virtual reconstruction of the artwork. To flesh out the presentation of the artwork, the images of the modeled artwork are encoded and decoded using Gaussian kernel functions, convolutional layers, and channel attention. The loss function is also optimized, and an end-to-end approach is used to generate the images. After analysis, it is concluded that the display characteristics of different artworks in the intelligent virtual simulation technology for artwork display are over 80%, among which all display characteristics of fresco works are over 90%, and the interaction degree as well as the humanity degree are also over 90%. Therefore, based on this technology, the artworks can be displayed well.