Fast Global Active Contour Model with Local Information
Liyan Wang, Jing Liu, Teng Wu · 2019
Based on the edgeless active contour (CV) model, an improved model is proposed in this paper. Because CV model only uses global intensity information to complete image segmentation process, so the processing of inhomogeneous images is not good, and even there will be error segmentation. In view of the disadvantage of CV model, we consider adding local gray information items to the energy function of CV model. In order to speed up the evolution of the model, we improve the global and local terms according to the method of reference [9]. At the same time, the energy penalty term is added to the energy function to avoid the re-initialization of CV model. Finally, in order to verify the practicability and validity of the improved model, some images are selected for MATLAB simulation experiments. The experimental results of the improved model are compared with those of the original CV model and the model in reference [8]. It is concluded that the improved model has higher segmentation accuracy and efficiency for uneven gray images.