Region-aided Geodesic Active Contour model for image segmentation

Abstract

In this paper, a novel algorithm for image segmentation is proposed. The presented method embeds the Geodesic Active Contour (GAC) model into the region based method. The proposed model adds a term that relates the region information of image to be segmented to the energy function of Geodesic Active Contour model. As a result, a new energy function is established. This method therefore includes both the edge and region information of the object instead of only the edge information as that in the original GAC model. Experimental results demonstrate that the proposed region aided GAC method is much more effective than the original one, especially when dealing with images with holes, weak edges and noises.