In this paper, we reconstruct the corresponding 3D face model using only one 2D image. 3D feature points are obtained by optimally approximating 2D feature points set with defined similarity. Then, a shape model is reconstructed by the warp function algorithm. Finally, a realistic face model is created through texture mapping with registration method such as affine transform. Results show that the models we reconstruct are comparatively realistic, and they can be used for face recognition or computer animation. The computation speed is also satisfying. © Springer-Verlag Berlin Heidelberg 2006.
CITATION STYLE
Lu, C., & Jie, Y. (2006). Automatic 3D face model reconstruction using one image. In Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics) (Vol. 4153 LNCS, pp. 235–243). Springer Verlag. https://doi.org/10.1007/11821045_25
Mendeley helps you to discover research relevant for your work.