3D Facial Expression-Synthesis Based on Feature Points Deformation Matching
Zhou Jian-hua · Journal of Xi'an University of Arts and Science · 2010
In allusion to the problem of inaccurate recovery from partly covered face image in video monitoring,the author puts forward a new algorithm of 3D facial expression synthesis based on feature points deformation matching.The triangular face network is set up after target face image graying.Some uncovered characteristics spots of the face image are extracted.A new network face model is synthesized after common model characteristic deformation matching.Expression texture is mapped separately from the front and the side.Three-dimensional facial expression images are synthesized according to illumination and aging parameters.The experimental results show that this approach is superior to the other similar algorithms of synthesis since the time complexity is lower,the accuracy is higher,and the robustness is stronger.Its advantage becomes more obvious with the increasing occultation of the characteristics.