Custom 3D virtual image generation system and method

Abstract

The invention discloses a user-defined 3D virtual image generation system. The system comprises a Maya binding module, a file management module, a classification feature analysis module, a face analysis generation module, an image resource generation module and a rendering interaction module. The invention further provides a user-defined 3D virtual image generation method which is based on Maya skeleton binding, BlendShape deformation, deep learning and feature preset parameter types and has a 3D virtual image generation capability with standard animation adaptability. And directivity adjustment can be carried out on the generation result according to subjective needs. A user binds a standard model of a 3D virtual character in Maya, loads a real person image through a file management module, analyzes the real person image in a generation module, classifies and identifies facial secondary features and facial auxiliary features through deep learning, and parameterizes facial features, and a rendering interaction module applies parameters to the 3D standard image. The 3D virtual anthropomorphic image of the specific animation is dynamically generated according to the role, and the generated result can be directionally adjusted according to subjective needs.

Description

technical field [0001] The invention relates to the field of 3D virtual anthropomorphic image generation, in particular to a self-defined 3D virtual image generation system and a generation method based on Maya binding automation combined with manual redirection. Background technique [0002] In the current software application field, digital twins, VR, AR, MR and image visualization are all visible future important product and technology trends. In these applications, setting roles with high image similarity and relevance for specific personnel can greatly improve user immersion, user participation, and product interactivity, and ultimately increase the added value of products; on the other hand, for More high-quality and high-efficiency generation of bound 3D avatars in line with correlation and similarity. With the continuous progress of deep learning in recent years, the technology of generating 3D images based on portrait photos is also gradually maturing. [0003] H...

AI Technical Summary

Problems solved by technology

[0005] In view of the above-mentioned deficiencies existing in the prior art, the purpose of the present invention is to provide a custom 3D virtual character generation system and generation method, which solves the problem that the 3D virtual anthropomorphic image of a specific animation cannot be dynamically generated according to the character, and the generated result cannot be adjusted according to subjective needs. Problems with directivity adjustment

Images