Fast implicit active contour models

Abstract

Implicit active contour models are widely used in image processing and computer vision tasks. Most implementations, however, are based on explicit updating schemes and are therefore of limited computational efficiency. In this paper, we present fast algorithms based on the semi-implicit additive operator splitting (AOS) scheme for both the geometric and the geodesic active contour model. Our experimental results with synthetic and real-world images demonstrate that one can gain a speed up by one order of magnitude compared to the widely used explicit time discretization. © Springer-Verlag Berlin Heidelberg 2002.