Virtual image model generation method, device, and storage medium

Abstract

The embodiment of the invention discloses a virtual image model generation method, a device and a storage medium, belonging to the computer technical field. The method comprises: A plurality of face images and a feature point matrix thereof are obtained, in accordance with that characteristic point matrix, the feature point matrix of natural human face image and the feature point matrix of non-natural human face image, Average matrix, identity shift matrix and expression shift matrix are obtained, facial identity base and facial expression base are obtained by dimension reduction processing, and a virtual image model matching facial expression with facial expression in target facial image is generated according to target facial image, facial identity base, facial expression base and virtual image model base. As only fivesensing organs are used for calculation, that time consumption is short, long-time occupation of processing resources is avoided, the operation performance is improved,and the data amount of the face identity base and the face expression base can be reduced by dimension reduction processing. Moreover, the virtual image model of expression matching can be generatedaccording to the facial identity base and facial expression base, which improves the versatility.

Description

technical field [0001] The embodiments of the present invention relate to the field of computer technology, and in particular to a method, device and storage medium for generating an avatar model. Background technique [0002] With the rapid development of multimedia technology and the continuous improvement of users' entertainment needs, a function of generating virtual image models has emerged as the times require. Through this function, facial expressions in human face images can be recognized, and facial expressions can be generated to match human facial expressions. avatar model, so as to be able to display their own expressions through the avatar model. This method is novel, unique and interesting, and has been welcomed by the majority of users. [0003] In related technologies, it is possible to shoot when the target user poses various expressions, obtain multiple face images, extract feature points from each face image, and obtain a feature point matrix of each face...

AI Technical Summary

Problems solved by technology

[0005] The coordinates of a large number of feature points are used to obtain the facial expression base, resulting in a large amount of data for the facial expression base, and the acquisition process takes a long time, requiring a long time to occupy processing resources and reducing operational performance

And the facial expression base is only applicable to the target user, not to other users, and is not universal

Images