Rapid 3D face reconstruction by fusion of SFS and Local Morphable Model

Various modeling methodologies have been proposed to model the human face realistically. However, despite the fact that these methods can give high-quality results, the time, cost, and labor required to generate such models by these methods are often high. In this paper, we propose an integrated modeling approach combining Shape-From-Shading (SFS) and Local Morphable Model (LMM), which can rapidly create a realistic 3D face model on the basis of a single 2D image rapidly. This method obtains 3D contour face quickly by SFS firstly; then, recovers local accurate 3D data by LMM; lastly, reconstructs a 3D face model of high-accuracy by carrying out interpolation smoothing on the 3D contour face with local 3D accurate data. Experimental results show that this method has the advantages of modeling with higher accuracy and lower complexity, and can reconstruct personalized 3D face model from a single real-world photo in a short time.

► 3D face modeling based on Shape-From-Shading is bad for face organs.
► Local Morphable Model perform better than Global Morphable Model in machine time and model accuracy.
► The method based on SFS and LMM is good for 3D face modeling.