A new image segmentation method based on weighted cut combination

The computation complexity of normalized cut image segmentation method based on graph theory is high, it is increasing with geometric multiples especially the vertex number of figure is more. Aimed at this shortage, this paper proposes a multi-scale level image segmentation method based on weighted cut combination. The method includes three parts, first of all, use the image pixel as vertex to structure the weighted figure, construct the cost function of weighted cut combination of undirected weighted figure; then, coarsening processing on the image, gain the coarse segmentation of highlighting interest area in undirected weighted figure; finally, the inverse interpolation method is adopted to improve the border combination for the highlighting interest area, get the optimal segmentation image. Experiments show that the method in this paper reduces the algorithm complexity compared with the classic normalized cut method, at the same time improves the segmentation accuracy and segmentation effect. Copyright © 2015 Binary Information Press.