Geometrical image segmentation by the Allen–Cahn equation Original Research Article

We present an algorithm of pattern recovery (image segmentation) based on the solution of the Allen–Cahn equation. The approach is usually understood as a regularization of the level-set motion by mean curvature where we impose a special forcing term which lets the initial level set closely surround the pattern in question. We show convergence of the numerical scheme and demonstrate function of the algorithm on several artificial as well as real examples.