Border Segmentation Using an Improved GGAC Model with Points Distance and Gray Intensity

An improved geodesic active contour model with points distance and gray intensity is proposed in this paper for the segmentation of images. Unlike the traditional geodesic active contour model, which has some disadvantages such as the selections of the contrast constant in the edge stop function, a new adaptive selection method has been developed to select the contrast constant and shrinkage velocity automatically. The proposed model extends the traditional geodesic active contour model and correctly segments objects with lower contrast. Experimental results show the effectiveness of the proposed algorithm.