Three-dimensional face reconstruction method and three-dimensional face reconstruction system for automatic multi-view-angle face auto-shooting image

Abstract

The invention discloses a three-dimensional face reconstruction method for an automatic multi-view-angle face auto-shooting image. The three-dimensional face reconstruction method comprises the following steps: automatically positioning mark points of a multi-view-angle face image of the same person; establishing a target function according to the positioned mark points and mark points corresponding to a reference face model to solve camera parameters; designing a reconstruction target function; and converting a three-dimensional face reconstruction problem into a multi-label image partitioning problem under a Markov random field, and solving by using a multi-label image partitioning algorithm. The method can be used for reconstructing a thick and precision three-dimensional face model and does not depend on an outer database, so that full-automatic face reconstruction can be realized and manual interaction does not need to be carried out by users.

Description

technical field [0001] The invention relates to the field of computer vision, in particular to a three-dimensional face reconstruction method and system for automatic multi-view face self-portrait images. Background technique [0002] Face reconstruction is one of the important research directions of 3D reconstruction. It has broad application prospects in the fields of film and television, games, 3D face recognition, etc., and is valued by researchers in the fields of computer graphics, computer vision, machine vision, and computer-aided design. . From the perspective of data acquisition, 3D face reconstruction is mainly divided into active ranging equipment and passive imaging equipment. Active ranging equipment such as laser scanners can scan and obtain accurate three-dimensional information of static objects. However, they are expensive, have a long scanning time, and have limited scanning range, making them difficult to use in applications that require high real-time p...

AI Technical Summary

Problems solved by technology

In addition, this type of face reconstruction method based on 3DMM needs to be combined with an aligned 3D face database, and the reconstruction result is obtained by linear superposition of the face database. The ability to detail the face

[0005] The deficiencies of the existing face reconstruction methods based on multi-view images are mainly as follows: 1) Due to the particularity of the sparse texture features of face images, the method based on traditional feature point matching is not applicable in practical applications, and it is difficult to use the traditional method based on The method of matching feature points to obtain dense 3D data; 2) requires cumbersome human interaction; 3) relies on an external 3D face database, and the accuracy of the reconstruction results depends on the richness of the database

Images