In this paper we have introduced a new method for image segmentation based on pairwise region comparison. We have shown that the notions of a segmentation being too coarse or too fine can be defined in terms of a function which measures the evidence for a boundary between a pair of regions. Our segmentation algorithm makes simple greedy decisions, and yet produces segmentation that obey the global properties of being not too coarse and not too fine according to a particular region comparison function. The method runs in O(mlogm) time for m graph edges and is also fast in practice, generally running in a fraction of a second.
Implementation of the segmentation algorithm described in:
Efficient Graph-Based Image Segmentation
Pedro F. Felzenszwalb and Daniel P. Huttenlocher
International Journal of Computer Vision, 59(2) September 2004.