Segmentation d'images par sélection de courbes de niveau

We study an image segmentation energy, which minimizer can be determined. The approach estimates the unknown number of objects and draws object boundaries by selecting the \best" level lines computed from level sets of the image. As a consequence, no energy minimization methods is necessary, yielding to a fast and non-iterative segmentation algorithm. Finally, anisotropic di usion is used to smooth level lines in noisy images.