Three-dimensional face reconstruction method, device and storage medium thereof

Abstract

The invention provides a three-dimensional face reconstruction method, a device and a storage medium thereof, which relate to the technical field of image processing. The three-dimensional face reconstruction method comprises the following steps: acquiring depth data of a target face under a plurality of viewing angles; judging whether the depth data under each view has a local depth data anomalyor not; if yes, re-acquiring the depth data under the perspective corresponding to the local depth data deformity point to update the depth data under the plurality of perspectives; obtaining a three-dimensional model of the target face according to the depth data in the plurality of views. This method collects depth data by depth camera when there are local depth data distortion points, which improves the error correction ability of face modeling and obtains more accurate three-dimensional face model.

Description

technical field [0001] The present invention relates to the technical field of image processing, in particular to a three-dimensional face reconstruction method, device and storage medium thereof. Background technique [0002] With the rapid development of computer equipment, network and image processing technology, the traditional naked eye image recognition method has been gradually replaced by the computer-automatic image recognition method, which greatly improves the efficiency and accuracy of image recognition. Based on the depth data of the face, automatic 3D face modeling by computer is a common task in the field of computer vision, and there are many applications. [0003] However, when the existing 3D face modeling method performs point cloud matching on the depth data and the face plane image, local depth data distortions often occur due to calculation errors, data collection errors, etc., and the local depth data distortions will be extremely The ground affects t...

AI Technical Summary

Problems solved by technology

[0003] However, when the existing 3D face modeling method performs point cloud matching on the depth data and the face plane image, local depth data distortions often occur due to calculation errors, data collection errors, etc., and the local depth data distortions will be extremely The Earth Affects the Accuracy of 3D Models of Faces

However, the existing technology requires a single scan of the depth data of the complete face at different angles before performing point cloud registration. It does not have real-time feedback logic and is prone to local depth data distortions, so it is difficult to effectively control the quality of depth data collection.

Images