Method, system and storage medium for displaying virtual images based on virtual-real combination technology

Abstract

The application relates to a method and system for displaying a virtual image based on a virtual-real combination technology and a storage medium, which comprises the following steps: collecting a two-dimensional target image containing a target object, inputting the two-dimensional target image into a preset prediction model to obtain human body three-dimensional prediction information, color prediction information, material prediction information and corresponding light prediction information of the target object, generating a three-dimensional reconstruction image under no light according to the three-dimensional prediction information, the color prediction information and the material prediction information, and rendering and processing the three-dimensional reconstruction image according to the light prediction information to obtain and display a three-dimensional reconstruction image with light. The model is trained through human body images under multiple different lights, the prediction result of the prediction model is more accurate, in addition, the light prediction information of the human body image is compensated based on the color prediction information of the human body image and the material prediction information of the human body image in the process of training the prediction model, the accuracy of the prediction result is further improved, and a more realistic three-dimensional reconstruction image with light is obtained.

Images