Comparison of Active Contour Acceleration Methods

J. Olivier, R. Boné, and J.-J. Rousselle (France)


Active contour, greedy algorithm, line search, shifted neighborhood, deformed neighborhood.


Active contours represent an important part in image processing and computer vision research for the past fifteen years. Unfortunately, the use of active contours in real-time applications is still not widely in used due to the slowness of the current algorithms. In this paper we are presenting several methods to accelerate active contours. We have developed three methods of optimization based on the greedy algorithm. The first one is based on anticipating the direction followed by each point, the sec ond and the third both on modifying the neighborhood of the studied point.

Important Links:

Go Back