Active contours (snakes) provide a unified account of a number of visual problems, including detection of edges, lines and subjective contours as well as motion tracking and stereo matching. Since their first apparition the researchers of the image processing community have simply applied this model to their problems without explaining where its parameters come from nor how to obtain its motion equations. To get a reality-based interpretation of the classic snake parameters (i.e. elasticity and bending) we propose an approach originates from the theory of the elasticity. Then we derive the motion equations from the variational method and we finally give some examples of segmentation. An alternative way of computing the external force field which increases the performance of snakes in presence of concavities is studied. We finally propose a new approach based on the dynamic modification of fi that improves the snake behavior at corners.