Integrated vector-valued active contour model for image segmentation

In this paper, image segmentation based on an integrated vector-valued active contour model is presented. Consider that each image channel has its signal characteristics, the region-based information uses the hybrid mean intensities simultaneously. Furthermore, by utilizing a two-dimensional vector field with different image channels, which provides different image patterns are used to constrain the results of image segmentation using edge-based information, edge-based information is also used, which extracts the information and uses the nonlinear function to learn the specific segmentation process. With the incorporation of the vector-based region and edge information, the proposed method can deal with multi-channel images effectively. The method is applicable for color images, multiresolution representation from frequency transformation, and multi-modal images, and can effectively overcome the problem that the weak edge of the image cannot converge due to large noise and poor contrast. It can confirm the effectiveness and robustness of the proposed method.