A convergence-to-boundary segmentation method combining GVF with GAC

In recent decades, image segmentation based on PDE is used widely in industry. Geodesic active contour (GAC) model is a common used method. But one drawback of this model is that it´s difficult to control the number of iterations and sometimes may produce an over-segmentation result. In this paper, a convergence-to-boundary method is proposed. In this method, the model combining gradient vector flow (GVF) with GAC is used when the distance between the evolution curves produced by successive iterations is smaller than some given threshold. It can extend the capture range and stop at the boundary stably avoiding over-segmentation. The experimental results show the curve can converge to boundary well.