Human face image warping based on local constraints

This paper proposes an algorithm of face image warping based on local constraints. Our method firstly specifies a limited number of feature points as the center of deformation, then finds their locations in transformed image and calculates their influencing radiuses respectively, after then calculates the geometric transformation of pixels within influencing areas according to local warping formula, finally calculates pixel gray values by means of bi-linear interpolation. Experimental results show that this approach is flexible, simple, fast and efficient.