Minimal surfaces: a geometric three dimensional segmentation approach

摘要: A novel geometric approach for three dimensional object segmentation is presented. The scheme based on deformable surfaces moving towards the objects to be detected. We show that this model related computation of minimal area (local surfaces). space where these are computed induced from image in which general also shows relation between classical obtained via energy minimization and ones derived curvature flows surface evolution framework. stable, robust, automatically handles changes topology during deformation. Results existence, uniqueness, stability, correctness solution presented as well. Based an efficient numerical algorithm evolution, we present a number examples detection real synthetic images.