Generalized Newton Method for Minimization of a Region-Based Active Contour Model

PED-based image segmentation based on the active contour model attracts many researchers due to the high precision of edge detection and the continuity of boundaries. Its basic idea is to define an energy functional on a dynamic curve which achieves its minimum when the curve conforms to the boundary of the objects. The most widely used optimization method is the gradient-descent method. However, the convergence of the gradient-descent method is very poor. In this paper, the effectiveness of the generalized Newton method is investigated by using it to minimize the energy functional of the RSF&CV model, which is a simple combination of the CV model and the RSF model. The experimental results show the accuracy and efficiency with robustness in noise.