Multi-view face synthesis method based on tensor resolution and Delaunay triangulation

Abstract

The invention discloses a multi-view face synthesis method based on tensor resolution and Delaunay triangulation, and mainly solves the problem of difficulty in face image synthesis under continuous changing views in the prior art. The method comprises the following steps of: extracting outline information of a face image by a characteristic point labeled method; separating out a view coefficientmatrix of human face data in a training set by a tensor resolution method; fitting sample bands of the view coefficient matrix; building characteristic points of a new view by adopting a tensor resolution formula; synthesizing the face image of the new view by adopting the Delaunay triangulation and linear affine transformation according to the known face image. The method has the advantages of authentic face synthesis result, wide synthesis view ranges and low operation complexity and can be applied to the field of computer vision or face synthesis under different views in a multimedia technology.

Description

technical field [0001] The invention belongs to the technical field of image processing, and in particular relates to a multi-view human face synthesis method, which can be used for the synthesis of human face images under different viewing angles in the field of image restoration and reconstruction. Background technique [0002] With the rapid development of IT technology, face-based identity authentication will become one of the most important contemporary biometric identification technologies. The face recognition test results jointly conducted by the US Defense Advanced Research Projects Agency DARPA and the US Army Research Laboratory ARL show that the face recognition rate has a strong sensitivity to attitude changes. Therefore, face synthesis technology from multiple perspectives is crucial to obtain reliable identity authentication. In addition, synthesizing face images under continuously changing perspectives is not only of great significance for face space modelin...

AI Technical Summary

Problems solved by technology

[0006] Since the above two types of methods do not accurately describe the nonlinear changes of face viewing angle factors, the first method can only synthesize face images under limited viewing angles, and due to the influence of multiple training image identities, the synthesized There is a large difference between the image and the real image; although the second type of method can obtain a better synthesis effect under the conditions of illumination and expression changes, due to the strong nonlinear changes of the multi-view angle of the face, the synthesis result is only close to the front Image from a limited view of a human face

Images