Abstract: On the basis of the existing 3D face reconstruction method, using pose estimation for depth information recovery was proposed, which is used for 3D face reconstruction based on a single image. It first obtained face pose parameters by matching the detected feature points to a general 3D face model, then applied pose parameters to the generic 3D face model to get depth information. Secondly, global and local deformation were made by detected feature points and their corresponding depth information to get the face geometry, then obtained texture coordinates by an as-similar-as-possible planar parameterization, finally got the 3D face model. Experiments have shown that using depth information recovered from pose estimation was efficient and easy to achieve good results.