A Bayesian Contour Measure for Image Segmentation

The Bayesian approach to image segmentation defines a penalty function of image partitions such that the function’s minima correspond to perceptually salient segments. We extend previous approaches following this framework by requiring that our image model sharply decrease in probability as a segment’s boundary is perturbed from its true position. Instead of making segment boundaries prefer image edges, we add a term to the penalty function that seeks abrupt change in the global mid–level representation and not just the local image brightness. We also introduce a prior on the shape of a salient contour that expresses the observed multi-scale distribution of contour curvature for physical contours. We show theoretically and experimentally that our new penalty term correlates strongly with salient structure. We apply our method to real images and verify that the new term improves performance. Comparisons with other state–of–the–art approaches validate our method’s advantages.