An MRF framework for co-solving image segmentation and border ownership

Image segmentation and border ownership assignment are two widely studied areas in the computer vision literature. It is well known that both the segmentation and the border ownership assignment play an important role in the visual perception. In this study, a Markov Random Fields model which provides a dual solution for the segmentation and the border ownership assignment is proposed. The proposed system is analyzed both quantitatively and qualitatively.