Information theory based snake adapted to inhomogeneous intensity variations
Résumé
This paper proposes a new snake based segmentation technique of a single object (simply connected) in presence of inhomogeneous Gaussian noise, for which the mean in each region is modelled as a polynomial function of the coordinates and which is thus adapted to inhomogeneous illumination. It is shown that the minimisation of the stochastic complexity of the image, which can be implemented efficiently, allows one not only to automatically estimate the number and the position of the nodes of the polygonal contour used to described the object, but also the degree of the polynomials which model the variations of the mean.