Controlling Growing Snakes by Using Key-Points

Olof Henricsson and Walter Neuenschwander Communication Technology Laboratory Swiss Federal Institute of Technology ETH CH-8092 Zurich, Switzerland Abstract In the present paper we address the problem of establishing contours in static monocular scenes. We propose to use features from uni ed edge and keypoint detection to drive an active contour model. We present a framework for image segmentation consisting of, (a) detection of 1D and 2D intensity discontinuities, i.e. general edges and key-points, (b) reformulation of the active contour model to an expanding version, called a \growing snake", (c) automatic initialization, (d) image and internal geometry driven expansion, and (e) controlled termination. The objective is to explore the potential of a fully automatic segmentation scheme. The performance and present limitations of growing snakes are demonstrated on images representative for diverse classes of scenes.