Image Foresting Transform: On-the-Fly Computation of Segmentation Boundaries

Abstract

Abstract. The Image Foresting Transform (IFT) is a framework for seeded image segmentation, based on the computation of minimal cost paths in a discrete representation of an image. In two recent publications, we have shown that the segmentations obtained by the IFT may be improved by refining the segmentation locally around the boundaries between segmented regions. Since these methods operate on a small subset of the image elements only, they may be implemented efficiently if the set of boundary elements is known. Here, we show that this set may be obtained on-the-fly, at virtually no additional cost, as a by-product of the IFT algorithm.

Joint graph cut and relative fuzzy connectedness image segmentation algorithm

“…In order to amend this problem, some authors consider the usage of regularization energies (e.g., internal forces) intrinsic to their formulations (Boykov et al, 2001; Malladi et al, 1995; Shi and Malik, 2000; Grady, 2006), while others enforce smoothness by post-processing (Malmberg et al, 2010, 2011; Malmberg, 2011; Falcão et al, 2002). The main drawback of the second group of methods is that it may be too late to fix a result, when it is already too far from the goal (e.g., significant leakage).…”