Procrustes — based shape prior for parametric active contours

Abstract

A novel method of parametric active contours with geometric shape prior is presented in this paper. The main idea of the method consists in minimizing an energy function that includes additional information on a shape reference called a prototype. Prior shape knowledge is introduced through a complete family of Euclidean invariants, computed from the similarity between shape of evolving contour and the prototype. This similarity is measured by full Procrustes distance. This extra knowledge enhances the model robustness to noise, occlusion and complex background. We use genetic algorithm to minimize energy function of this new type of snake that we call it Procrustes snake. The variational formulation of the proposed approach is described in details. We obtain promising results with synthetic and real images which show the power of our method for segmentation tasks.