A method of creating 3-D face images from 2-D photos for face recognition
Xin Guan, Hanqi Zhuang · International Journal of Biometrics · 2010
In this paper, a method of creating a Three-Dimensional (3-D) face image from one or more 2-D face images is proposed. The 3-D synthetic face is actually represented by parameters of a set of 3-D face images of other persons. When a 2-D probing photo is presented, a 3-D synthetic face image is projected onto the 2-D image plane, following the pose and illumination compensations. An optimisation algorithm is applied to minimise the error, once the algorithm converges, the depth and intensity parameters of the 3-D synthetic face image are obtained of the 2-D probing photo in the 3-D space.