The Visual Computer, vol. 12, no 2, 1996, pp 75-83.

A practical solution for tracking edges in image sequences with snakes

Michael Hoch* , Peter C. Litwinowicz

Apple Computer Inc.
ATG-Graphics
1 Infinite Loop, MS 301-3J
Cupertino, California 95014

*) Academy of Media Arts
Dept. of Computer Graphics
Peter-Welter-Platz 2, 50676 Cologne, Germany
email: micha@khm.de

Abstract
Active contour models, or "snakes," developed in (Kass et al. 1988), use a simple physical model to track edges in image sequences. Snakes as originally defined however, tend to shrink, stretch and slide back and forth in unwanted ways along a tracked edge and are also confused by multiple edges, always grabbing the nearest one. In this paper a practical solution is presented that combines motion estimation techniques with snakes to overcome these problems. An algorithm is presented that uses a block matching technique to guide the endpoints of the snake, optical flow to push the snake in the direction of the underlying motion, followed by the traditional snake edge-fitting minimization process. We use this technique for tracking facial features of an actor for driving computer animated characters.

Key words: Pattern Recognition - Edge Tracking - Active Contour Models - Facial Animation

	showing the moving endpoint problem, i.e. see the endpoints of the snake move away from the mouth edge position specified in the first frame
	shows the the snake switching edges due to too much motion, i.e. after two frames the upper edge of the lower lip is tracked
	applying the proposed technique using block matching on the snake endpoints and optical flow to acount for motion the correct lip edge is tracked.