An improved RBF deforming algorithm in 3D face morphable model

An improved RBF (Radial Basis Function) deforming algorithm, which is based on the RBF interpolation of scattered data and aims at the deficiency of RBF algorithm in 3D face morphable model, is proposed in this paper. According to the different characteristics of feature points and non-feature points, it chooses different parameter values for the basis function to make the adjustment transformation more reasonable in the model. Then it executes a secondary smoothing on the local areas containing feature points again by use of the binary cubic polynomial interpolation method to avoid producing the distortion model, and further characterizes the specific feature details of the face organs. Ultimately it obtains a smooth and natural personalized 3D face model.