Method, device and electronic equipment for building a three-dimensional facial model

Abstract

The invention provides a three-dimensional face model establishing method and device and electronic equipment, belongs to the technical field of computers, relates to artificial intelligence and machine learning technologies, and performs three-dimensional object reconstruction based on the machine learning technology and the computer vision technology in the artificial intelligence. The three-dimensional face model establishing method comprises the steps: establishing a basic three-dimensional face model according to an acquired face image; obtaining a depth increment graph of the basic three-dimensional face model relative to the face image; and projecting the depth increment graph to the surface of the basic three-dimensional face model to obtain a detail three-dimensional face model. According to the embodiment of the invention, a basic three-dimensional face model is established for an acquired face image; and then obtaining a depth increment graph of the basic three-dimensional face model relative to the face image, and projecting the depth increment graph to the surface of the basic three-dimensional face model to obtain a detail three-dimensional face model, so that the precision of three-dimensional reconstruction can be improved, and a more realistic and more natural detail three-dimensional face model can be obtained.

Description

technical field [0001] The present application relates to the field of computer technology, and more specifically, to a method, device and electronic equipment for establishing a three-dimensional face model. Background technique [0002] With the rapid development of computer technology, 3D reconstruction technology has been widely used in the field of computer vision technology. As an important branch of 3D reconstruction technology, 3D face reconstruction is widely used in film and television animation, game entertainment, video surveillance and other fields. For example, in the game industry, a large number of human characters and anthropomorphic game characters have been created since the three-dimensionalization of games. With the improvement of hardware level, many virtual reality and augmented reality games and applications have emerged. In these games and applications, various characters are becoming more and more vivid and real, so high-precision 3D face models ar...

AI Technical Summary

Problems solved by technology

Means modeling generally requires experienced professionals to use Autodesk, 3DMax and other software to complete, but it requires a lot of manpower and time costs

The 3D data modeling based on instrument collection is mainly based on the 3D data collected by precision 3D acquisition instruments such as depth cameras, structured light or laser scanners for 3D face reconstruction. This method is based on the real 3D data of objects for modeling, with high precision. However, the equipment used is generally expensive, and the use requires professional training, so its application market is limited

Relatively speaking, the cost of image-based automatic modeling is low. The existing image-based automatic modeling methods generally use the general face model as the basis, and deform or deform the general face model according to the facial key point information in the image. Adjust to obtain the 3D model corresponding to the face in the image. The accuracy of 3D reconstruction using this method is low, and it is impossible to obtain a realistic and natural 3D face model.

Images