Consideration is given to the use of snakes (active contour models) for tracking objects in a visual field. The basic theory of snakes is covered, and a greater depth of analysis is given to both the original  and the Fast Snake  implementations. The strengths and limitations of these models, for tracking deformable objects using unaided visual systems, are discussed. Suggestions are made that their future lies in their integration with neural network visual location prediction systems.
|Name||UH Computer Science Technical Report|
|Publisher||University of Hertfordshire|