Image modeling and region-based active contours segmentation

In recent years, the field of active contour-based image segmentation has seen the emergence of two competing approaches : Boundary-based and region-based approaches. After recall of these two aproaches, we present an algorithm for objects segmentation using region-based active contours. From the minimization of a general criterion, we propose a new Eulerian method to compute the evolution equation of the active contours. The computed evolution speed expression makes the active contour evolve as fast as possible towards a minimum of the criterion. This Eulerian method allows to easily take in account the information attached to the regions, information that varies with the evolution of the contour. Experimental results on real images are presented.