Study of shape from shading algorithm based on the radial basis function network
基于径向基函数网络的SFS算法研究

After analyzing the shortcomings of some methods on smoothness and continuity for Shape From Shading(SFS),a new SFS algorithm based on the radial basis function model was proposed.Firstly it constructed a curved face equation,and then did self-learning by adjusting power gene and center and width of the radial basis function with restriction condition on reflect function,finally gained satisfactory curved face equation.Theoretical analysis and experimental results show that this algorithm improves a lot on veracity of shape recovering and smoothness and continuity of the curved face.