Merging Active Contours
In current active contour image segmentation methods, the number of regions is assumed to be known beforehand. It is related directly to a fixed number of active curves. How to allow it to vary is an important question which has been generally avoided. This study investigates a segmentation prior related to regions area to allow the number of regions to vary automatically during curve evolution, thereby optimizing the objective functional implicitly with respect to the number of regions. The obtained evolution equations show that the proposed prior can cause some curves to disappear while other curves expand, thereby leading to a region merging by curve evolution, although not in the sense of the traditional one-step merging of two regions. We give a statistical interpretation to the coefficient of this prior to balance its effect systematically against the other functional terms. We show the validity and efficiency of the method by testing on real images of intensity. A comparison demonstrates the advantages of the proposed method over the region-competition algorithm in regard to the optimal number of regions and computational load.
KeywordsActive Contour Initial Number Real Image Curve Evolution Data Term
Unable to display preview. Download preview PDF.
- 10.Kadir, T., Brady, M.: Unsupervised non-parametric region segmentation using level sets. In: Proc. Int. Conf. on Computer Vision, pp. 1267–1274 (2003)Google Scholar