Bio-holographic image segmentation by using interactive graph-cut

Digital holographic microscopy has been studied for bio-applications. In order to analyze specimens in the holographic image, image segmentation as a preprocessing step is inevitable. The main energy-based segmentation includes snake, level-set and graph-cut. For the snake and level-set, initial curve is needed to be added by the user. However, for the different initial curves, different segmentation result will be obtained. Graph-cut based segmentation can obtain the global optimal result. Furthermore, graph-cut method utilizes both regional and edge information which can make the segmentation more reasonable. In this paper, in order to reduce the effect of the discontinuity of the object edge for the segmentation and ensure the segmentation result perfect, a few of user interferences are applied.