3D Reconstruction From A Single Image Based On Shape Evolution
Qian Ma · Advances in computer science research · 2015
3D reconstruction from a single image is a challenging problem.In this paper, we present an improved method for 3D reconstruction.We use improved arrival time method to solve the Hamilton-Jacobi equation which is obtained from the method of shape from shading (SFS).The concave-convex ambiguity problems occurred like with other SFS methods, so we find out the boundary of the concave-convex regions according to 3D Segmentation based on Gaussian curvature, then reconstruct the surface use the shape evolution method based on energy constraint in order to obtain the accuracy 3D shape.