A new alternating minimization algorithm for image segmentation

A new alternating minimization algorithm for image segmentation is proposed. Based on the similarities between Ambrosio-Tortorelli model and Perona-Malik model, a new explicit alternating minimization algorithm is developed. The Euler-Lagrange equations of Ambrosio-Tortorelli model is modified to produce sharper edges. The effect of parameters is studied. Numerical results are presented to illustrate the effectiveness of the proposed model.